ERP-based detection of brain pathology in rat models for ...€¦ · Figure 9: ERP feature...
147
ERP-based detection of brain pathology in rat models for preclinical Alzheimer’s disease by Seyed Berdia Nouriziabari A thesis submitted in conformity with the requirements for the degree of Master of Science Department of Cell and Systems Biology University of Toronto © Copyright by Seyed Berdia Nouriziabari 2016
Transcript of ERP-based detection of brain pathology in rat models for ...€¦ · Figure 9: ERP feature...
ERP-based detection of brain pathology in rat models for preclinical Alzheimer’s disease
by
Seyed Berdia Nouriziabari
A thesis submitted in conformity with the requirements for the degree of Master of Science
Department of Cell and Systems Biology University of Toronto
© Copyright by Seyed Berdia Nouriziabari 2016
ii
ERP-based detection of brain pathology in rat models for preclinical Alzheimer’s disease
Seyed Berdia Nouriziabari
Master of Science
Department of Cell and Systems Biology
University of Toronto
2016
Abstract
Early pathological features of Alzheimer’s disease (AD) include the accumulation of
hyperphosphorylated tau protein (HP-tau) in the entorhinal cortex and progressive loss of basal
forebrain (BF) cholinergic neurons. These pathologies are known to remain asymptomatic for
many years before AD is clinically diagnosed; however, they may induce aberrant brain processing
which can be captured as an abnormality in event-related potentials (ERPs). Here, we examined
cortical ERPs while a differential associative learning paradigm was applied to adult male rats with
entorhinal HP-tau, pharmacological blockade of muscarinic acetylcholine receptors, or both
conditions. Despite no impairment in differential associative and reversal learning, each
pathological feature induced distinct abnormality in cortical ERPs to an extent that was sufficient
for machine classifiers to accurately detect a specific type of pathology based on these ERP
features. These results highlight a potential use of ERPs during differential associative learning as
a biomarker for asymptomatic AD pathology.
iii
Acknowledgments
“If I have seen further it is only by standing on the shoulders of giants.”
-Issac Newton (1642-1727)
My first debt of gratitude must go to my advisor Dr. Kaori Takehara-Nishiuchi. I would like to
thank her for her invaluable mentorship, patience and support; and for giving me the space and
time to grow and improve. Special thanks to my committee Dr. Blake Richards and Dr. Liang
Zhang for their support, guidance, and helpful suggestions. Their guidance served me well and I
owe them my heartfelt appreciation.
A special acknowledgement also goes to past and current members of Takehara Lab who have
richly colored the last few years of my studies and who have provided me with much needed
technical and moral support, as well as lively discussion. In no particular order, I would like to
thank Stephanie Tanninen, Dr. Julien Volle, Dr. Mark Morrissey, Dr. Nathan Insel, Maryna Pilkiw,
XiaoTian Yu, and Susmita Sarkar.
To all my friends, thank you for your understanding and support in my many, many moments of
crisis. Your friendship makes my life a wonderful experience. I cannot list all the names here, but
you are always on my mind.
I dedicate this manuscript to my parents and my dear Julie, who have always been my foundation
and a constant source of encouragement. Although it may not be apparent to many people; I would
have not been able to complete this work without them.
This thesis is only a beginning of my journey.
iv
Table of Contents
Abstract ........................................................................................................................................... ii
Acknowledgments.......................................................................................................................... iii
Table of Contents ........................................................................................................................... iv
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
Chapter 1 Introduction .......................................................................................1
1.1 The history and background .............................................................................................1
1.2 Theories on etiology of Alzheimer’s disease ....................................................................4
1.2.1 Amyloid theory .........................................................................................................................4
1.2.2 Tau theory .................................................................................................................................5
1.2.3 Cholinergic hypothesis .............................................................................................................7
1.3 Alzheimer’s disease types ..................................................................................................8
1.3.1 Familial AD & genetic mutations.............................................................................................8
1.3.2 Sporadic AD .............................................................................................................................9
1.4 Stages of Alzheimer’s disease ............................................................................................9
1.4.1 Dementia stage .........................................................................................................................9
1.4.2 Mild cognitive impairment stage ............................................................................................10
1.4.3 Preclinical stage ......................................................................................................................11
1.5 Current biomarkers of Alzheimer’s disease ..................................................................15
1.5.1 Genetic biomarkers .................................................................................................................15
1.5.2 Cerebrospinal fluid biomarkers ..............................................................................................16
1.5.3 Blood biomarkers ...................................................................................................................17
1.5.4 Structural MRI ........................................................................................................................18
v
1.5.5 Cerebral blood flow and metabolism .....................................................................................19
1.5.6 Molecular imaging..................................................................................................................20
1.5.7 Electroencephalogram ............................................................................................................21
1.5.8 Event-related potentials ..........................................................................................................22
1.6 Limitations of current ERP biomarkers ........................................................................26
1.7 Strategies to improve ERP biomarkers .........................................................................28
1.8 Hypothesis and Approach ...............................................................................................30
Chapter 2 Methods ........................................................................................... 31
2.1 Subjects .............................................................................................................................31
2.2 Design ................................................................................................................................31
2.3 Surgeries ...........................................................................................................................31
2.4 Viral vector infusion ........................................................................................................32
2.5 EMG-EEG electrode Implantation ................................................................................32
2.6 Drug treatment .................................................................................................................33
2.7 Apparatus and Behavioral paradigm.............................................................................33
2.8 Behavioral testing coupled with EEG recording ..........................................................34
2.9 EMG Analysis...................................................................................................................34
2.10 EEG analysis....................................................................................................................35
2.11 Support Vector Machine ................................................................................................36
2.12 Statistical analysis ...........................................................................................................37
2.13 Immunohistochemistry ...................................................................................................38
Chapter 3 Results .............................................................................................. 45
3.1 Entorhinal hyperphosphorylated tau over-expression .................................................45
3.2 Behavior ............................................................................................................................51
3.2.1 Acquisition .............................................................................................................................51
3.2.2 Reversal 52
vi
3.3 Event-related potentials...................................................................................................55
3.3.1 Auditory stimulus-evoked ERPs in the frontal cortex ............................................................55
3.3.2 Auditory stimulus-evoked ERPs in the parietal cortex ..........................................................56
3.3.3 Auditory stimulus-evoked ERPs in the temporal cortex ........................................................57
3.3.4 Visual stimulus-evoked ERPs in the frontal cortex ................................................................58
3.3.5 Visual stimulus-evoked ERPs in the parietal cortex ..............................................................59
3.3.6 Visual stimulus-evoked ERPs in the temporal cortex ............................................................60
3.4 SVM-based classification of ERPs ..................................................................................67
Chapter 4 Discussion ........................................................................................ 72
4.1 Comparison of histological changes with AD staging ...................................................72
4.2 Effects of preclinical AD pathology on associative learning and reversal learning ...72
4.3 Mechanisms underlying ERP abnormality and susceptibility of different cortical
regions to features of preclinical AD pathology ............................................................73
4.3.1 HP-tau-induced synaptic dysfunction and cortical hyperexcitability .....................................74
4.3.2 Cholinergic deficiency and the hippocampus .........................................................................75
4.3.3 HP-tau, cholinergic deficiency, and the abnormal frontal ERPs ............................................76
4.4 Use of ERP features as a detection tool for early brain pathology ..............................77
4.4.1 Effect of stimulus modality and contingency on ERP-based detection of pathology ............78
4.4.2 Susceptibility of ERP properties to preclinical AD pathology ...............................................79
4.4.3 Detection of AD pathology with a single ERP component ....................................................80
4.5 Summary and conclusion ................................................................................................81
4.6 Limitations ........................................................................................................................81
4.7 Future Direction ...............................................................................................................82
References ......................................................................................................................................85
vii
List of Tables
Table 1: Coordinates of the viral vector infusion
viii
List of Figures
Figure 1: Pathological hallmarks of Alzheimer's disease
Figure 2: Progression of pathological features in Alzheimer's disease
Figure 3: From EEG to ERPs
Figure 4: A cartoon representation of rAAV9 vector
Figure 5: Bi-lateral infusion of the viral vector into the entorhinal cortex
Figure 6: EEG and EMG electrode placements
Figure 7: Differential trace eyeblink paradigm
Figure 8: Time windows for capturing ERP peaks after CS onset
Figure 9: ERP feature extraction and data set combinations for SVM
Figure 10: Summary of steps in SVM classification
Figure 11: Expression of hyperphosphorylated tau in the entorhinal cortex
Figure 12: DTEBC Acquisition (10 Days)
Figure 13: DTEBC Reversal (7 Days)
Figure 14: Auditory ERPs during DTEBC paradigm
Figure 15: Interaction graphs of auditory ERP amplitude
Figure 16: Interaction graphs of auditory ERP latency
Figure 17: Visual ERPs during DTEBC paradigm
Figure 18: Interaction graphs of visual ERP amplitude
Figure 19: Interaction graphs of visual ERP latency
ix
Figure 20: SVM classifier accuracy
Figure 21: Summary SVM classifier accuracy
1
Chapter 1 Introduction
1.1 The history and background
In 1910, Emil Kraepelin published the 8th edition of the book Psychiatrie in which he included a
description of work done by his student, Alois Alzheimer (1). The name Alzheimer’s disease (AD)
was coined by Kraepelin for description of symptoms and pathology of the disease. In his work
Alzheimer observed memory impairment and cognitive worsening in his 55 year old patient,
Auguste Deter, and described the presence of neurofibrillary tangles (NFTs) and senile plaques
(SPs) in a postmortem sample of her brain (2).
Today, AD is the most common and well-known form of dementia affecting the increasingly
elderly population. Included in AD symptoms are deterioration of memory and cognitive functions
that progress over 10 to 15 years and are associated with the formation of SPs comprised of
amyloid beta (Aβ), accumulation of NFTs consisting of hyperphosphorylated tau (HP-tau), and
progressive atrophy of cholinergic neurons in the basal forebrain (BF) (Figure 1) (3-8).
Currently, over 750,000 Canadians 65 and older are suffering from AD and the direct and indirect
costs of AD reaches a total of $10.4 billion per year (9). Despite recent advancements in AD
therapeutics, it remains as one of the most challenging diseases to tackle. This is due to the long
asymptomatic course of progression of AD which cannot be detected by current primary screening
and diagnostic criteria largely rely on conventional tests on memory and executive function in at-
risk individuals.
To put things into perspective, we can compare AD to an uncontrollable train that is speeding
towards the end of a railway track. The onset of the first pathological process in the brain starts
when the motionless train begins to move. At first, it goes unnoticed that the train is gaining speed
until it approaches the end of the track. The time that has elapsed up until this moment represents
the preclinical stage of the disease where cognitive symptoms are limited in spite of extensive
brain pathology (10). As pathology becomes more severe, it ultimately reach a threshold where
cognitive symptoms become evident. Returning to the train journey, we manage to board the
2
speeding train in hopes of preventing the imminent crash by pulling the emergency break, only to
discover that the break has been damaged and the crash is inevitable. Similarly, once AD is
diagnosed it is too late for it to be cured or slowed down, since the pathology has already caused
severe damage to the brain. The remaining options for patients after AD diagnosis are to make
arrangements for assisted living and use treatments that may ease a subset of disease symptoms
(11).
To change the series of event surrounding the train we must notice the train as soon as it begins to
move and apply the emergency break before it becomes broken. In AD these actions respectively
translate into preclinical diagnosis prior to development of extensive and irreversible brain
damage, and development of disease-modifying or arresting treatments. Hence, it is imperative to
detect AD in its preclinical stage where there might still be a chance to halt or significantly delay
the advancement of the disease.
3
Figure 1. A) A neuron displaying early pathological hallmarks of Alzheimer’s disease.
Accumulation of intracellular neurofibrillary tangles (NFTs; red) comprised of
hyperphosphorylated tau protein, and formation of extracellular senile plaques (SPs; purple)
of amyloid beta aggregates.
Figure 1. Pathological hallmarks of Alzheimer's disease
Figure 1. B) Projections of basal forebrain cholinergic neurons from the basal nucleus (Bas)
and medial septal nucleus (MS) to several different cortical and subcortical regions in a rat
brain (Blue; Entorhinal cortex, EC; Basolateral amygdala, BLA). Through the course of
Alzheimer’s disease, degeneration of basal forebrain cholinergic neurons results in significant
reduction in cholinergic input and disruption of cholinergic modulation at the efferent brain
region. Anatomical regions named according to the rat brain atlas (Paxinos G, Watson C.,
2007).
NFTs
SPs
Axon
A
B
Bas MS
4
1.2 Theories on etiology of Alzheimer’s disease
There has been several theories about the causes of AD since the first time it was described by
Alois Alzheimer in 1907 (12). Once a correlation was found between the numbers of SPs, NFTs,
and cognitive decline in 1960’s, researchers began probing the underlying mechanisms involved
in formation of these pathologies (13). Moreover, AD attracted immense amount of research when
it was officially listed as a neurological disorder in 1970’s and was no longer considered a part of
normal aging (13).
Majority of current theories on AD pathogenesis emphasize the presence of excessive deposition
of abnormal proteins found in SPs and NFTs, such as Aβ peptide and HP-tau. However, imbalance
of brain’s homeostatic environment in addition to bacterial and viral agents that are able to cross
the blood-brain-barrier have also been proposed as possible causes of AD (14-16) Here, the
foremost theories on the causes of AD pathogenesis are presented.
1.2.1 Amyloid theory
One of the theories for the cause of AD involves accumulation of Aβ protein. Although much is
not known about the function of Aβ protein, mutation in the gene of its precursor protein, amyloid
precursor protein (APP), has been implicated in familial AD. APP is suggested to be important for
neural growth (17-19) neural signaling, and may also function as a metalloprotein to mediate
copper transport and metabolism (20, 21).
In most cell types APP undergoes the non-amyloidogenic pathway resulting in the production of
the P3 peptide fragment, which consists of 16 amino acids and involves α-secretase cleavage
followed by a γ-secretase cut within the Aβ domain of the APP protein (28, 29). The amyloidogenic
pathway involves cleavage by β-secretase followed by the γ-secretase, releasing the 40-43 amino
acid Aβ peptides which are secreted extracellularly and are the main component of amyloid
plaques in AD (28-30). This longer form (Aβ42) is more hydrophobic and fibrillates more easily,
and is also thought to be more neurotoxic than the Aβ40 peptide (28, 29). These peptides can be
found as monomers, oligomers, and fibrils and in the latter stages form the SPs seen in the brains
of AD patients.
5
It is speculated that aggregation of SPs as well as the soluble form of Aβ oligomers can have
detrimental effects (31-33). This is supported by studies indicating the neurotoxicity of Aβ
oligomers against neurons using transgenic mice that over-expressed APP (33). It has also been
suggested that SPs may be brain’s defense mechanism that store and curb neurotoxic oligomers
(34). Nonetheless, age-dependent elevation in cortical Aβ peptide level is attributed to decreased
effectiveness of mechanisms responsible for its removal, such as phagocytosis by microglia and
astrocytes, and enzymatic degradation by neprilysin and insulin-degrading enzyme (35).
Moreover, studies in rats have reported that Aβ oligomers inhibits long term potentiation (LTP) in
the dentate gyrus of hippocampus (36, 37), and that they attenuate the neuronal excitability in
mouse dentate gyrus by reducing the number of action potentials in response to current stimulation
(38). Findings from other studies also suggest that accumulation of Aβ oligomers decreases the
level of excitatory transmission at the level of synapse, and induces epileptiform discharges and
abnormal pattern of neuronal circuit activity at the network level (39, 40).
The distribution of SP deposits in the brain initially starts in the neocortex and subsequently
expands to the diencephalic nuclei, striatum and cholinergic nuclei in the BF, brainstem, and
ultimately the cerebellum. Secondary to SP accumulations, NFT formation and neuronal
dysfunction occur which provokes clinical symptoms of AD (43). Recently, the validity of the
amyloid theory has been challenged as newly developed vaccines successfully removed SP
deposits but failed to remove NFT or enhance patients’ cognition (44). The common occurrence
of SP lesions in non-demented elderlies also sheds doubt upon their involvement in causing AD
(45).
1.2.2 Tau theory
Due to the weak correlation between SPs and neuronal death in AD, researches begin to argue that
maybe the other hallmark of AD pathology, HP-tau is responsible for the observed neuronal death
in AD patients. Tau protein supports tubulin assembly into microtubules which provide structure
and flexibility to the cell (46-48). Tau is ubiquitously found in the central nervous system and
particularly in neurons. Alternative splicing and phosphorylation of tau messenger ribonucleic acid
(mRNA) produces six different isoforms of the protein which are all expressed from MAPT gene
6
on chromosome 17 (48, 49). These isoforms range from 441-452 amino acids and contain different
number of binding domains, but all are present in AD (47-49). Tau phosphorylation and de-
phosphorylation is a normal homeostatic process, however this mechanism is shifted towards tau
hyper-phosphorylation in AD (48).
Disequilibrium in tau phosphorylation results in detachment of HP-tau from the microtubules and
its aggregation into compact paired helical filaments (PHF) inside the cell. This results in
disruption of microtubule assembly and ultimately leads to formation of NFTs and cell death
observed in postmortem AD brains (47). After the death of the cell, NFTs remain as ghost tangles
and can be observed by silver staining or immunohistochemical (IHC) techniques (50).
In addition to HP-tau as the main constituent of NFTs, heparin sulphate, fibroblast growth factor,
casein kinase II, proteoglycan, ubiquitin, and microtubule association protein are also present in
NFTs (22-25). It has been suggested that these proteins may act as a secondary mechanism for
formation of brain lesions in AD (22-25, 50).
Braak and Braak classified AD progression into six stages (Braak I-VI) based on the site and
density of NFT (3). In stages I-II (transentorhinal) NFTs are mostly confined to the entorhinal and
perirhinal regions. Transentorhinal stage corresponds to the asymptomatic preclinical stage of AD
(54). Stages III-IV (limbic) present with elevated incidence of NFTs in limbic structures, where
the numbers of NFTs in hippocampus correlate with the severity of cognitive symptoms. Lastly,
spread of NFTs into the isocortical regions completes AD progression in stages V and VI (54).
HP-tau and NFTs are also observed in other neurodegenerative disorders collectively known as
tauopathies, including Frontotemporal dementia, Pick’s disease, sporadic corticobasal
degeneration, argyrophilic grain disease, and supranuclear palsy (48, 64, 65). Studies have linked
hereditary frontotemporal dementia and Parkinsonism to mutations in tau MAPT gene on
chromosome 17 in absence of SP (66). Occurrence of neurodegenerative disorders such as
hereditary frontotemporal dementia in patients with mutated tau and in absence of SP indicates
that mutation in tau protein alone is sufficient to create a toxic form of tau that can cause damage
and neuronal death.
Imbalance of activity between kinases and phosphatases responsible for tau modulation has also
been proposed to contribute to NFT formation (31). It is an accepted theory that formation of HP-
7
tau occurs after Aβ accumulation has already begun (47, 48). However, due to its well-observed
neurotoxicity and predictive ability of dementia severity, some suggest abnormal tau to be the
initiator of neurodegeneration (47, 48, 67). Nonetheless, the exact mechanisms behind these
lesions are still to be determined, including whether they are causative in AD, or a secondary
events to some other underlying and as yet unknown pathology.
1.2.3 Cholinergic hypothesis
The cholinergic hypothesis was theorized based on the compelling evidence that showed BF
cholinergic neurons, which send numerous projections to cortical and limbic structures, degenerate
and result in a lower acetylcholine (ACh) level in AD patients compared to non-demented elderlies
(68-70). It is suggested that degeneration may be a result of NFTs which were observed inside the
cell body of the BF cholinergic neurons (71-77).
Implication of cholinergic system in AD was substantiated by post-mortem neurochemical analysis
of brain tissue obtained from AD patients. Labelling of the enzyme, choline acetyltransferase
(ChAT), responsible for synthesis of acetylcholine and specific to cholinergic neurons, showed a
significant reduction (60-90 %) in ChAT activity in the cerebral cortex and hippocampal formation
of patients who died of AD compared to age-matched non-demented elderlies. Conversely,
analysis of muscarinic acetylcholine receptor (mAChR) distribution showed no difference in the
density of the receptors between AD brains compared to healthy controls. Therefore, the
cholinergic deficiency in AD is thought to be largely due to lack of presynaptic input rather than
loss of postsynaptic receptors (73, 78).
Post-mortem studies provided the initial evidence for involvement of cholinergic system in AD
neuropathology. Further evidence was provided by studies which demonstrated that suppression
of cholinergic activity in normal individuals results in impairment of complex behavioral tasks
(79, 80). It has been long known that a transient amnesic state could be induced by central mAChRs
blockers (83). Administration of low doses of scopolamine, a non-competitive mAChRs blocker,
to young adult volunteers resulted in impairment of recent memory but did not affect working or
long term memories (84). Moreover, it is known that administration of central cholinergic
potentiators improves formation of recent memory and negates the effect of mAChR blockers (85).
8
Together, these findings suggested that the impairment of short-term memory in early-stage AD
is caused by disruption in cholinergic input and HP-tau induced hippocampal neural death (86-
90).
Thus, AD treatments (donepezil, rivastigmine, and tacrine etc.) focused on increasing the levels of
ACh in the brain by inhibiting cholinesterase, the enzyme responsible for ACh hydrolysis.
However, long-term clinical assessments showed that cholinesterase inhibitors provide limited
symptomatic treatment without any disease modifying effects (91). Limited effects of
anticholinesterase drugs on AD symptoms plus the distressing side effects demonstrate the need
for development of more efficacious therapies for patients with AD (92).
1.3 Alzheimer’s disease types
Dominant familial AD accounts for less than 1% of diagnosed patients (93). Familial AD is caused
by genetic mutations and has an early manifestation of symptoms before the age of 65 (1, 94).
Sporadic AD has a much higher prevalence, however the causes of its pathogenesis and the risks
pertaining to the disease are still unclear.
1.3.1 Familial AD & genetic mutations
Much information has been gathered from familial AD about the pathogenesis of the disease in
both familial and sporadic forms. Studies on inflicted individuals with familial AD revealed that
mutations in either APP gene (APP, chromosome 21) or the genes of the proteins responsible for
APP cleavage, Presenilin 1 (PSEN1, chromosome 14) and Presenilin 2 (PSEN2, chromosome 1)
can cause elevated production of the longer isoform of Aβ peptide (95-97). This isoform may then
aggregate to form SP which is proposed to disrupt neuronal signaling and ultimately result in
neuronal death (99).
9
1.3.2 Sporadic AD
Sporadic AD accounts for the majority AD diagnosis. Currently, there are no definitive causes for
sporadic AD, however it is speculated that multiple factors may be involved in instigating the
dysfunction that leads into the manifestation of symptoms recognized as AD (100, 101). Recently,
differentiation between AD subtypes has been proposed to be important for predicting disease
progression and assigning corresponding treatments (102). It has been observed that different
cases of the AD may present with variable proportions of AD pathological hallmarks, suggesting
the existence of a SP-dominant and NFT-dominant AD subtypes (103-105). By incorporating
different variables such as, cerebrospinal fluid (CSF) levels of tau and Aβ peptide, time of disease
onset, and the number of apolipoprotein 4 (APOε4) alleles on chromosome 19, it has been
suggested that AD may have up to five subgroups (102). As in many complex disorder, several
factors such as genetic, dietary, and environmental may trigger AD pathogenesis and incite its
progression.
1.4 Stages of Alzheimer’s disease
By incorporating recent scientific and technological advances, the new guideline for diagnostic
criteria divides AD into three stages, including 1) dementia due to Alzheimer’s, 2) mild cognitive
impairment (MCI), and 3) preclinical Alzheimer’s. Dementia due to AD is characterized by
dysfunction of memory, thinking, and behavior which significantly interferes with individual’s
functioning in everyday life (106). In MCI stage, deficits in memory and behavior are not severe
enough to interfere with individual’s daily functions in life, however these deficits are observable
and can be measured with mental status tests (107). Current findings suggest that in the preclinical
stage detectable changes in brain may occur many years before manifestation of any impairment
in memory, thinking, or behavior (108).
1.4.1 Dementia stage
Patients in this stage show apparent memory deficits, detected as disruption in episodic memory
which allows an individual to consciously retrieve a previously experienced item or episode of life
(109). In early AD, NFTs accumulate in the EC and hippocampus whereas Aβ plaques are first
10
observed in the neocortex (110-118). It is in these areas that the subsequent synapse dysfunction,
neuronal loss and brain atrophy is first observed. (119-126). At the time of the clinical AD
diagnosis, there is a substantial level of neuronal loss in the medial temporal lobes (MTL), and the
location of damage appears to reflect the type cognitive impairments present (122-25) The damage
spreads to the parietal and frontal lobes sequentially as the disease progresses (112, 127).
Cognitive disturbances in AD are required for clinical diagnosis and can be classified in several
groups. (119, 128, 129). The degree of cognitive impairment is an important indicator of disease
severity and can be utilized as a predictor and a risk factor for advancement of MCI to AD.
Difficulties with delayed recall as well as episodic memory are two most common symptoms of
AD at the time of diagnosis (76, 119, 130-132).
The new criteria revision of National Institute of Neurological and Communicative Disorders and
Stroke and the Alzheimer's disease and Related Disorders Association (NINCDS-ADRDA),
suggests isolated disturbances in episodic memory a sufficient symptom for AD diagnosis (94).
Moreover, based on the findings of large community cohort studies, early stage AD can be
precisely differentiated from healthy state by task that examine episodic memory, perceptual
speed, semantic memory, executive and visuospatial functions. (133-136).
Besides memory impairment at the initial stage of the disease, other classical cognitive symptoms
include, agnosia, apraxia, and aphasia (119, 130). In patients with advanced AD, there is an evident
worsening of initial symptoms and emergence of frontal lobe dysfunction which manifests as
change in personality, mood fluctuation, and apathy (130). In late AD, majority of cognitive
abilities are deteriorated and the patient is unable to communicate with the surrounding
environment (130). In addition to cognitive deficits, depression is also commonly present in
individuals with AD at the time of diagnosis (130). Overall, the sequence in which AD
neuropathology advances in the brain dictates the order of the manifested cognitive deficits (112,
137).
1.4.2 Mild cognitive impairment stage
Mild cognitive impairment (MCI) is a heterogeneous condition that has been linked to an increased
risk of developing AD. Neuropathological features of MCI resemble AD neuropathology with a
11
substantial amount of NFT and Aβ plaques (10, 104, 117, 121, 138-141). The prevalence of AD
neuropathology in individuals with MCI is between 75-95%, with amnesic MCI having the highest
prevalence (141, 142).
Consistent with previous studies, the number of NFTs in the MTL and the degree of neuronal
death, particularly in the layer II of the EC, is substantially observed and is closely relate to
memory dysfunction in MCI (115,140). Additionally, it has been reported that the intensity of
neuropathology in MCI, including the degree of neuronal loss, the number of Aβ plaques in the
neocortex, and NFTs in the MTL is at an intermediate level between age-matched controls and
early-stage AD. (122, 125, 140). Overall, the pathological features of AD, as observed in MCI,
precede clinical AD diagnosis and are almost fully developed before or concurrent with the first
objective cognitive impairment.
Moreover, MCI is defined by combination of impairment in cognitive abilities (143). Due to
heterogeneous nature of MCI, the subset of affected cognitive functions are different among
individuals; however problems with episodic memory appears to be a consistent trend in MCI and
have a strong predictive value for conversion to AD (131, 137, 128, 134-137). Disturbance in
perceptual speed and learning are another indicators of MCI with progressive type and strong
predictors for AD development (131, 135, 137, 143, 146, 148).
Compared to individuals who display impairment only in one cognitive function, subjects with
deficits in multiple cognitive domain appear be at an increased risk of MCI conversion to AD (135,
148-151). Further, as several cognitive deficits are found in both stable and progressive type of
MCI, additional biomarkers are needed to reliably distinguish individuals with stable MCI from
subjects with progressive MCI (107).
1.4.3 Preclinical stage
Numerous studies have reported the presence of NFT and Aβ plaques in the brain of elderlies who
did not exhibit any cognitive impairment (10, 104,, 113-116, 138, 139, 141, 152-161). Although
neuronal loss in asymptomatic elderlies is minimal in contrast to individuals with MCI and AD,
pathological changes in the brain follow the progression of AD neuropathology and first affect the
MTL and neocortex with NFT and Aβ plaques respectively (10, 111, 113-117, 121).
12
The prevalence of AD neuropathology reported in asymptomatic elderlies varies between studies
from 10% to 40%. This variability is mostly likely due to the differences in classification criteria
or difficulties establishing a standardized definition for non-demented elderlies. (88, 104, 108, 109,
116). Nonetheless, the current consensus regards elderlies with neuropathological hallmarks of
AD in absence of cognitive symptoms as preclinical AD (10, 113, 104, 139, 140). The different
types AD neuropathological features and their proposed levels in preclinical, MCI, and AD are
depicted in figure 2 (162).
Overall, although neuropathological findings allow for differentiating between healthy versus AD
state, as well as determining the stage of the disease, they have limited applications in clinical
settings since obtaining a brain autopsy of live individuals for neuropathological assessment is not
possible. Therefore, biomarkers other than histological features are needed to detect
neuropathological changes associated with AD and possibly identify individuals at the preclinical
stage.
13
Path
olo
gic
al
Norm
al
Preclinical MCI AD
Figure 2. The proposed advancement of different pathological features during the course of
Alzheimer's disease (AD) progression (Jack et al., 2010). In the preclinical stage, accumulation of
Amyloid-β (purple) is observed at the neocortical regions. Synaptic dysfunction (green),
hyperphosphorylated (HP)-tau-mediated neuronal injury (blue), and atrophy of basal forebrain
cholinergic neurons (pink), and atrophy of cortical and subcortical structures (brown) appear to begin
later during the preclinical stage. In the mild cognitive impairment stage (MCI), different pathological
features of AD continue to progress and first signs of cognitive dysfunction are detected (red). In early-
stage AD the severity of pathologies and the amount of damage is beyond repair.
Figure 2. Progression of pathological features in Alzheimer's disease.
Amyloid-β accumulation
Synaptic dysfunction
HP-Tau-mediated neuronal injury
Cholinergic deficiency
Brain structure atrophy
Cognition
14
Understanding of the preclinical phase of AD has important theoretical and clinical implications.
From a theoretical perspective, it may elucidate the mechanisms involved in disease progression
by expanding the knowledge on how normal aging may transition to dementia. Clinically, it may
allow for early identification of at-risk individuals and maximizing the efficacy of AD treatments
(163). The cognitive traits of preclinical AD have been investigated in several large longitudinal
community cohort of healthy elderlies, with longer than 20 years follow-up period in the longest
study (164). The findings indicated that multiple cognitive features, such as abstract reasoning,
attention, cognitive speed, and executive functions are affected in preclinical stage of AD, and
effective traits at identifying at-risk individuals (131, 135, 149, 164- 172).
Deficits of multiple cognitive functions validate the observation that multiple brain regions are
affected long prior to AD diagnosis. In particular, these results are in line with the knowledge that
disturbance in episodic memory is a persistent feature of AD, and a result of damage to regions of
MTL prior to manifestation of observable memory impairment clinical diagnosis (147, 164, 172-
174). Nonetheless, it is important to note that deficits in these cognitive domains are not unique to
AD, and that cognitive speed, executive functioning, and episodic memory also decline through
the course of adult lifespan (175-177).
Therefore, it is important to differentiate between cognitive disturbances associated with
preclinical AD and non-progressive cognitive deficits observed in normal aging. This
differentiation may be possible by identifying precipitating risk factors including but not limited
to, unrecognized hypertension, proneness to distress, head trauma, and stroke (178-181).
Furthermore, simultaneous examination of other functioning parameters may be useful for
detection of preclinical AD in at-risk elderlies. These include volumetric and functional
measurement of brain structures, deposition of Aβ plaques and HP-tau, glucose metabolism and
cerebral blood flow, subjective memory complaints and depressive symptoms, and genetic risk
factors such as apolipoprotein-4 allele (117, 174, 181-193).
15
1.5 Current biomarkers of Alzheimer’s disease
A biomarker is generally defined as a characteristic that can be measured accurately and
reproducibly, and that detects a fundamental and preferably unique feature of a neuropathology in
a specific disorder (8, 194). The purpose of a biomarker may be for diagnosis, monitoring disease
development, or prognosis and prediction the course of the disease (195). Moreover, a biomarker
can be categorized as a marker of trait, state, or rate (194). A trait biomarker indicates the
probability or susceptibility of developing a specific disorder. A state biomarker is used for
diagnosis, and a rate biomarker indexes the progression or the stage of the pathology (8, 194).
In the past decade neuropathological changes in AD have been regarded as the best standard
clinical criteria for diagnosis of probable AD, while the other AD biomarkers in molecular biology
and brain imaging have been mainly used within the research field (119). Recently, a working
group from the National Institute on Aging proposed that, an ideal biomarker for AD should have
sensitivity and specificity of at least 80% for detecting AD and differentiating it from other
dementias, it should be reliable, reproducible, non-invasive, inexpensive, and simple to perform
(8). However, based on a recent systemic review from the Alzheimer's disease and Related
Disorders Association (ADRDA), the best validated biomarkers that are currently being used for
early diagnosis of AD are genetic biomarkers, cerebrospinal fluid (CSF) biomarkers, structural
imaging with magnetic resonance imaging (MRI), and functional imaging with positron emission
tomography (PET) (196).
1.5.1 Genetic biomarkers
Thus far, most successful AD diagnostic exams rely on advanced techniques in molecular genetics,
and are limited to diagnosing familial AD with early onset. As discussed previously, mutations in
PS1, PS2, and APP genes cause dominant familial AD. Overall, these mutations have only been
identified in a few hundred families, in which PS1 mutation accounts for the majority of early-
onset AD below the age of 50, while APP mutations are rare, and PS2 mutations have only been
observed in two pedigrees (197-200). It is thus reasonable to look for mutations of PS1 gene in
early-onset familial AD; however, it is impractical to search for mutations in individuals with no
history of AD in the family.
16
Further, in contrast to deterministic genetic causes of AD, a specific apolipoprotein E (apoE) allele
is considered a risk factor for AD and has been shown to have 3 times more prevalence in late
onset sporadic AD compared to healthy controls. (197, 199). The apoE protein is a part of
intermediate-density lipoproteins that transports cholesterol to neurons via the apoE receptors
(200). There are three variants of apoE alleles on chromosome 19 found in general population,
including ε2, ε3, and ε4, which create apoE proteins that differ by a single amino acid residue
substitution, ε2 (Cysteine112, Cysteine158), ε3 (Cysteine112, Arginine 158), and ε4 (Arginine 112,
Arginine 158) (201). Among these allele variants, the ε4 subtype has been associated with increased
probability of both early-onset and late-onset AD of sporadic and familial type (202-204).
However, since apoE ε4 allele is only considered a risk factor, it is insufficient to provide a
diagnosis for AD solely based on genotyping of the apoE gene, as this measurement alone has a
low levels of sensitivity and specificity. Further, the majority of the individuals with inherited
apoE ε4 allele exhibit no sign of AD even at an old age. Nonetheless, a slight increase in diagnostic
confidence is obtained when apoE ε4 allele is found in parallel to conventional diagnostic markers
during clinical assessment. (200, 205).
1.5.2 Cerebrospinal fluid biomarkers
Much focus has been given to three proteins in the cerebrospinal fluid (CSF) as the potential
biomarkers for AD, including Amyloid beta 42 (Aβ42), total tau (t-tau), and phosphorylated tau (p-
tau). There is a general consensus that low levels of Aβ42 and high levels of tau in the CSF indicate
the presence of AD neuropathology as observed in post-mortem brains (208).
A comparison between AD patients and age-matched controls showed high levels of sensitivity
and specificity for t-tau and p-tau (81% sensitivity, 91% specificity) and Aβ42 (86% sensitivity,
89% specificity) (209). In addition, the specificity of an AD diagnosis can be improved by using
CSF biomarkers in conjunction with other diagnostic techniques, such as computed tomography
(CT) or MRI (210, 211).
Reports have shown that CSF biomarkers may also be able to predict MCI conversion to AD.
Convergent findings of five studies showed that MCI progression to AD was predicted by high
levels of t-tau and p-tau and low levels of Aβ42 in the CSF (212). However, a consensus is yet to
17
be reached over which of these three proteins or their combination is the strongest predictor for
MCI conversion to AD (213-216).
Overall, although CSF biomarkers may provide an accurate diagnosis for AD, obtaining CSF via
lumbar puncture is a procedure that is practiced only in specialized centers and requires
tremendous changes to services and healthcare systems in order to be implemented universally
(217). Further, sampling of CSF suffers from a negative public reputation, and is accompanied by
high rates of reservation among patients mainly due to its highly invasive nature and some
associated side-effects such as headaches and nausea (218).
1.5.3 Blood biomarkers
In contrast to CSF, detection of biomarkers through venipuncture provides a less invasive test. The
efforts to identify a reliable blood biomarkers for AD have not been very successful. One of the
challenges for finding a suitable blood biomarker is the low expression level of protein of interest,
such as Aβ, in the blood and the peripheral tissue which can challenge the protein quantification
assays (219). Moreover, compared to CSF, plasma and serum contain a much higher level of
proteins which can further hinder the detection of target proteins (219, 220).
There has been incongruent findings in regards to plasma levels of Aβ42. A few studies have
reported slight increase of Aβ42 in AD patients compared to healthy controls, while the majority of
studies did not show any change (221). Further, a few studies have suggested that high levels of
plasma Aβ42 indicates elevated risk of AD development, while findings from other studies did not
find evidence supporting this claim (221).
Unlike the negative correlation between CSF Aβ42 levels and cortical Aβ-plaques, as supported by
imaging studies, the correlation between plasma Aβ42 levels and Aβ-plaque is yet to be determined
(222). Findings in regard to plasma tau levels are also conflicting. Several studies have reported
an increase of tau levels in AD patients, while others observed a decrease of plasma tau levels in
the inflicted individuals (223, 192). The contributing factor to conflicting observations is likely
the extremely low levels of plasma tau which can result in variability between measurements.
Currently, ultrasensitive assays for quantification of plasma tau are being developed to address
this issue and examine the usefulness of plasma tau as a biomarker for AD (224).
18
Recently, advanced molecular approaches have characterized a few novel plasma biomarkers.
These novel biomarkers include C-reactive protein, selective anti-peptoid antibodies, and other
plasma proteins including IL 1alpha, IL-3, and TNF-alpha which can distinguish AD from healthy
controls (225-228). Nonetheless, these novel blood biomarkers require to be further studied before
being implemented as fully reliable biomarkers for AD diagnosis.
1.5.4 Structural MRI
Studies have shown that progression of AD neuropathology and the severity of cognitive deficits
have a strong correlation with the level of brain atrophy measured with structural MRI biomarkers
(228, 229). At the early stages of AD, degeneration begins in the structures of the MTL, including
the hippocampus and EC, and continues with an average annual rate of 4.7% compared to 1.4% in
age-matched controls (230).
Wide implementation of structural MRI analysis in clinical settings has been hindered mainly due
to its expensive setup and requirement of an on-site technical expert for measuring the volume of
the structure of interest based on the analyzed volumetric and visual tracing data (231). However,
recently the latter issues has been addressed by using automated data acquisition and analysis
technique, known as tensor-based morphometry, which measures global brain atrophy and
calculates group differences between the patients and controls (231).
To further enhance the accuracy of region-of-interest and global brain atrophy measurements,
supervised learning algorithms, such as support vector machine (SVM), are trained to
automatically classify MRI scans as AD or control, and produce a Structural Abnormality Index
(STAND) based on the pattern and level of atrophy observed in each voxel (232). Based on the
results of a multi-center study, STAND index can differentiate 98 AD patients from 109 controls
with an accuracy of 84% (232). In another study the classification accuracy was increased to 94%
for AD versus controls, when MRI and CSF biomarkers were combined (233).
Additionally, structural MRI biomarkers have also been able to predict MCI progression to AD
(205). It has been shown that MTL atrophy, measured using automated morphometry, combined
with cognitive performance scores can predict MCI progression to AD with 85.2% sensitivity and
90% specificity (234, 235).
19
Currently, structural MRI biomarkers provide an accurate non-invasive methods for diagnosing
AD patients. Nevertheless, this method of diagnosis is unusable for individuals with
claustrophobia, metal implants, or pacemakers. In addition, this method is not ideal for detection
of preclinical AD, as neuronal loss and structural atrophy during this stage is insufficient to
differentiate between preclinical AD and age-matched controls.
1.5.5 Cerebral blood flow and metabolism
Examination of AD with functional imaging techniques became rapidly explored after the
development of radio tracers for PET, such as fluorodeoxyglucose (18F-FDG), and tracer ligands
for single-photon emission computed tomography (SPECT), including hexamethyl-
propyleneamine oxime (99mTc-HMPAO), N-isopropyl-(iodine-123)p-iodoamphetamine (123I-
IMP), and ethylene cysteinate dimer (99mTc-ECD) (236, 237).
As a glucose analog, FDG captures spatial information about the neural metabolism of cortical and
subcortical structures by accumulating in tissues in proportion to their metabolic activity 30
minutes after injection. SPECT radio tracers provide information about cerebral blood flow by
binding to tissues with higher level of perfusion (238-240).
In early stages of AD, reduction of cortical metabolism and perfusion first occurs in the MTL, the
posterior cingulate, and the bilateral posterior temporo-parietal junction. As the disease progresses,
reduced metabolic activity and hypoperfusion are observed in the prefrontal cortex which are
thought to be associated with emotional disturbance and personality changes observed in later
stages of AD (241-253).
There appears to be an increased likelihood of conversion to AD in MCI patient who show a
decreased level of cortical metabolism in temporo-parietal association areas, posterior cingulate,
and hippocampus (254-258). Indeed, decreased metabolic activity in the EC and hippocampus
measure with PET have been used for discriminating control subjects from individuals with MCI
and AD with 81% and 100 % accuracy respectively (259).
A post-mortem analysis on the brains of AD patients who have been evaluated by SPECT revealed
a 84% sensitivity and a 92% specificity for this technique (260). Findings of a meta-analysis on
20
27 studies with HMPAO have indicated sensitivity of 77% and a specificity of 89% for this SPECT
ligand in discriminating AD patients from the age-matched controls (261).
The evidence suggest that both PET and SPCET are able to differentiate AD from other types of
dementia (262), and that a diagnostic index for AD can be provided from the ratio of association
cortex activity to the activity of primary sensory-motor cortex (263). However, the accuracy of
both techniques for diagnosis of AD can be further enhanced by analyzing imaging results in
combination with information gathered from clinical assessment (264).
1.5.6 Molecular imaging
Development of radio tracers for Aβ-plaques and tau aggregates allowed for in-vivo localization
of pathology and advanced the use of positron emission tomography (PET) in the study of AD.
Currently, 2-(1-(6-[(2-[F-18] fluoroethyl) (methyl) amino]-2-naphthyl) ethylidene) malononitrile
(18FFDDNP) and 11C-labelled Pittsburgh Compound B (PiB) have been most extensively studied
(207, 208). In comparison to conventional PET tracers such as, 18F-FDG, that indirectly measured
tissue metabolism, direct binding of 18FFDDNP to Aβ-plaques and NFTs provides an index which
predicts MCI conversion to AD (267, 268).
Consistent with the presence of NFTs in MTL in early-stage AD, the strongest binding of
18FFDDNP is also observed in this region, and performance on memory tests was positively
correlated with the rate of 18FFDDNP clearance from the brain (269). However, detection of NFT
pathology with this molecular tracers has proven challenging as the difference in signal intensity
of 18FFDDNP between AD patients and age-matched controls is not more than 10% (231).
In contrast, PiB has a higher sensitivity than 18FFDDNP for binding to Aβ-plaques and provides
close to 100% change in signal intensity between AD patients and controls (270, 271). However,
results from a study have shown patterns of PiB binding in 21% of controls that were not
distinguishable from the binding patterns in AD patients (272).
It remains unknown whether comparable PiB binding pattern in controls and AD patients was a
reflection of preclinical AD in the control subjects, or simply due to insufficient specificity of PiB
21
binding. Nonetheless, the PiB binding pattern in approximately 50% of MCI patients is similar to
AD and predicts MCI conversion to AD within a two-year period (273).
Unfortunately, PiB radio tracer has short half-life of 20 minutes, rendering PiB-PET-based study
of AD invasive and expensive. As such, wide implementation of this diagnostic technology in the
clinical setting for large-scale screening of at-risk individuals is yet unattainable (274, 275).
Currently, other Aβ-specific ligands, such as 18F-BAY94-9172 and 11C-BF-227, are being
investigated to establish their clinical relevance in AD (95, 96, 231). The advantages of these radio
tracers include longer half-life of 18F-BAY94-9172 (i.e. 110 minutes), and higher specificity of
11C-BF-227 than PiB to dense Aβ plaques (96, 96).
In addition, recent studies with a new radio tracer specific for tau, known as T807 or 18F-AV-1451,
have shown that there is a correlation between the level of tau radio tracer binding in the MTL, the
level of tau CSF biomarkers (i.e. t-tau and p-tau), and cognitive performance in AD patients (164,
276).
1.5.7 Electroencephalogram
Electroencephalogram (EEG) is the recording of the brain’s spontaneous electrical activity that
occurs because of synchronous firing of a large number of neurons. In a clinical setting EEG
recording is often quantified (qEEG) by decomposing it into different frequency bands and each
band is assessed for its rhythmic activity (304). The commonly used frequency bands are the delta
(0-4 Hz), theta (4-7 Hz), alpha (8-12 Hz), beta (12-30 Hz), and gamma (> 30 Hz) (304).
The characteristic EEG abnormalities in AD include shift of the power spectrum to lower
frequency bands and attenuation of coherence in high frequency bands (278). These abnormalities
are correlated with disease severity and are thought to be due to functional disconnections among
different cortical areas that result from synaptic dysfunction, axonal pathology, cholinergic
deficits, and neuronal death (279-281).
The shift of power spectrum to lower frequencies, or the slowing of EEG, has been extensively
observed in patients with MCI and AD (282-285). This change includes attenuation of power in
22
high frequency alpha and beta bands and increase of power in low frequency delta and theta band.
However, elevated activity in the gamma range has also been reported in some AD patients (286).
In early stages of AD, it is believed that the first changes in awake resting EEG include an increase
in theta power and decrease in beta power, and in later stages of the disease a decrease in alpha
power and an increase in delta power is observed (287-290).
Moreover, it has been reported that theta power in in early-stage AD patients is significantly
different compared to age-matched controls, but not compared to patients with MCI (287, 291,
292). Studies have also suggested that an increase in theta power is predictive of cognitive decline
in individuals with subjective memory complaint (291).
Pharmacological studies have also reported that administration of scopolamine to subjects induced
changes in EEG similar to what is observed in AD patients, and that these changes were eliminated
by administration of acetylcholinesterase inhibitor (293-296). Also, it has been shown that there is
a correlation between the activity of low-frequency EEG bands in AD and several other biomarkers
such as hippocampal atrophy in structural MRI (324), elevated CSF tau levels (298), and cortical
hypo-metabolism and hypo-perfusion (298-304).
1.5.8 Event-related potentials
Another way to analyze EEG data is event-related potentials (ERPs). ERPs are a series of negative
and positive voltage deflections of the EEG that are time-locked to sensory or cognitive events
(Figure 3) and have been validated as markers of AD in an outpatient setting clinical trial (305).
The first component of EPRs is a positive voltage deflection known as P1. P1 occurs approximately
50 ms after the onset of an auditory stimulus or 100 ms after a visual stimulus. Auditory P1
component has been associated with auditory inhibition and sensory gating (306, 307). Further,
studies using magnetoencephalography (MEG) have suggested the sources of auditory P1
component to be located in the superior temporal gyrus and medial prefrontal cortex (mPFC) (308,
309).
23
Figure 3. From EEG to ERPs.
Figure 3. To obtain event-related potentials (ERPs) from electroencephalogram (EEG), a time-
window on the EEG is specified before and after the stimulus presentation (red), and EEG in that
specific time-window is averaged across all trials. ERP waveform displays positive (P) and negative
(N) deflections known as P1, N1, P2, N2, and P3 ERP components arising in hundreds of
milliseconds (ms) after the onset of stimulus at zero (Luck et al., 2000).
Trial 1
Trial 2
Trial N
ERPs
Time (ms)
EEG
Averaged
24
Visual P1 has been proposed to reflect several sensory processing, such as noise suppression,
encoding of form and color, and the arousal level (299, 310, 311). Based on the data gathered from
PET, brain electrical signal analysis (BETA), and low-resolution brain electromagnetic
tomography (LORETA) analysis, multiple brain regions have been proposed to generate visual P1,
including ventrolateral occipital cortex, striate, and posterior fusiform gyrus (300-302).
The second ERP component, N1, is a negative voltage deflection that occurs approximately 70-
100 ms after the onset of an auditory stimulus and 100-120 ms after the onset of a visual stimulus
(303, 304). Studies have associated auditory N1 component with encoding of sensory and physical
properties of the stimulus, such as location and intensity, and have shown that the N1 amplitude is
increased by elevated levels of attention to the stimulus (312-316).
Results from MEG, lesion studies, and BESA analysis have indicated the primary auditory cortex
in superior temporal gyrus and additional structures in the frontal cortex as the likely generators
of auditory N1 component (317-321). It has been suggested that visually-induced N1 component
reflects preparation of response, as this component was attenuated in tasks that a motor response
was not needed (299). Further, it was shown that the amplitude of visual N1 is generally larger
during stimulus discrimination, likely due to elevated processing of the location and spatial
properties of the stimulus (299, 310, 321, 322). Using MEG, MRI, and LORETA, the sources of
visual N1 component have been located in the inferior occipital lobe, occipito-temporal junction,
as well as the inferior temporal lobe (323, 324).
The positive deflection after N1 ERP component is known as P2. The amplitude of auditory P2
can be observed as a single or a double peak 150-275 ms after the stimulus onset, and is suggested
to be sensitive to physical features of the stimulus, such as pitch and loudness (325-327).
According to MEG studies and intracranial recordings, it has been proposed that auditory P2 is
generated from sources in the primary and secondary auditory cortices (328-330). The amplitude
of visual P2 appears 150-200 ms after the stimulus onset over the frontal region and around 200
ms over the occipital areas (331-333). Further, more complex visual stimuli result in a larger P2
amplitude.
Overall, both auditory and visual P2 are thought to reflect cognitive tasks, including feature
detection, stimulus classification, and selective attention (334-339). The N2 ERP component is a
negative deflection that peaks approximately 200 ms after stimulus onset, and has been attributed
25
to multiple cognitive processes including stimulus discrimination, target selection, and mismatch
detection (340-343).
The results of scalp current density analysis suggest that bilateral sources in the auditory cortex
are responsible for generation of auditory N2 component, with the largest amplitude over the
central parietal area (263, 264). Intracranial recordings have localized the source of visual N2 at
the fourth occipital gyrus, and the largest visual N2 amplitude was recorded bilaterally close to
occipito-temporal sulci (344, 347).
The P3 component is the most extensively researched ERP component since its discovery in 1965
(353). It is believed that P3 reflects the activity of brain during higher cognitive processes, such as
attention and memory (354). This finding is based on the work of earlier studies in which stimulus
information such as, probability and task relevance, were manipulated in an “oddball” paradigm
to examine the effect on the pattern of brain electrophysiological activity (355-358). The results
from these studies showed that P3 is elicited in response to an unexpected auditory or visual
stimulus approximately 300 ms after stimulus onset (357).
It has been theorized that P3 elicitation to an unexpected stimulus is an index of brain’s attention-
dependent reassessment of the mental representation that was created by the stimulus in the
working memory (358-361). If no change in stimulus characteristic is detected, the mental
representation remains unchanged and only the earlier ERP components are observed (i.e. P1, N1,
P2). Upon detection of a change in stimulus attributes, the mental representation of the stimulus is
updated via attention-dependent mechanisms and a P3 deflection is observed (362). This context-
updating theory of P3 was based on the observations that discrimination of a rare target stimulus
from a frequent non-target stimulus elicits a prominent P3 deflection, and that the amplitude of P3
increases as the probability of the target stimulus presentation is lowered (363-365).
The two subcomponents of P3 are P3a and P3b potentials. P3a is observed when a non-task-related
rare stimulus, also known as a distractor, is detected among a series of frequents stimuli, whereas
P3b is produced upon detection of a task-related target stimulus (366, 367). The P3a displays a
rapid habituation and reduction of amplitude with repeated presentation of the distractor stimulus,
and therefore it is thought to act as an orienting response that is generated by communication
between the hippocampus and the frontal cortex (368-370).
26
The precise neural origin of the P3 component is yet unknown. In patients with a frontal lobe
lesions a reduction in P3a but not P3b amplitude was observed (369). Similarly, patients with focal
hippocampal lesions demonstrated an attenuated P3a but an unchanged P3b, indicating the
necessity of hippocampus and frontal cortex for generation of P3a potential (371). Further, results
from human studies using in-depth electrodes in hippocampus, along with assessment of patients
with severe damage to medial temporal lobe demonstrated only a partial involvement of the medial
temporal lobe in generation of the P3b potential (372, 373).
1.6 Limitations of current ERP biomarkers
Among biomarkers reviewed above, ERP biomarker holds a number of advantage over other
biomarkers, such as non-invasiveness, low cost, portability, and availability. Unlike blood, CSF,
and PET biomarkers, obtaining ERPs does not require the use of invasive procedures or exposure
to radiation. Also, the cost of required hardware for acquisition and recording is substantially lower
than majority of other biomarkers (375, 376). Compared to imaging biomarkers which have
temporal resolution of seconds, the millisecond temporal resolution of ERPs allows for
examination of processes underlying cognitive function and dysfunction (374). Additionally, ERPs
are more tolerant to movement than neuroimaging techniques, and there are methods that can
reduce or completely eliminate the movement artifact prior to analysis of the ERP data (377, 378).
Therefore ERPs are suitable for studies where subjects are required to perform a task, as well as
in behavioral studies with unrestraint animals.
In parallel, current ERP biomarkers also have some limitations. Many studies have reported
abnormalities in the amplitude and latency of P3 in MCI, AD patients (379-382) and elderlies who
are likely to be in preclinical stages (381, 383). However, observations have been conflcting, and
there is no general consensus over which P3 propetry is the the best candidate for detection of AD
at an early stage.
One study reported that an increase in frontal and parietal P3 latency, induced during an auditory
oddball paradigm, was correlated with a 3.7 fold increase in probability of AD in individuals with
subjective memory complaints (381). In another study, marginal reduction of frontal and parietal
27
P3 amplitude, induced during an auditory oddball paradigm, was observced in subjects at risk of
developing MCI without any abnormality in P3 latency (383).
Although these findings suggest a potential use of P3 as a biomarker for AD (384-387), there
appears to be large variability among subjects which undermines the use of P3 as the method of
assessment for clinical diagnosis or evaluation of at risk individuals (389).
Moreover, the abnormality of P3 is not always detectable on an individual basis, or between
subjects with distinct neurological disorders (384). A study that investigated P3 using an oddball
auditory paradim in patients with AD and individuals suffering from depression, reported that
although P3 latency was sufficient to detect the group difference between demented, depressed,
and age-matched controls, it was unable to differentiate a demented patient from the control due
to lack of sensitivity (389). In a similar study, P3 potential was examined using a two-tone auditory
oddball paradigm in individuals with AD of mild to moderate severity, and in patients with
depression (390). It was shown that, although AD gorup had the longest averaged P3 latecy in all
the scalp regions analyzed, no significant group difference was found between the AD, depressed,
and control group. Further, when P3 latency was used for distinguishing individual AD patients
from control the classificaiton accuracy was less that 25% (390).
The early ERP components, including P1, N1, and P2 are known as the obligatory cortical
potentials due to their strong within-subject and inter-individual consistency (348-350). As seen
in above examples, P3 component lacks in consistency and reliability (351, 352). The limitations
of this component may go beyond its inter and intra-individual variability. It has been suggested
that different stimulus modalities elicit P3 potentials with dissimilar characteristics. Several studies
in normal healthy adults reported that auditory P3 potentials display a shorter peak latency,
whereas visual P3 potentials show larger peak amplitude (391-393). Further, in patients with mild
to moderate AD, auditory and visual P3 components induced using an oddball paradigm display a
significant modality-specific alterations, including prolongation of averaged auditory P3 latency
and attenuation of averaged visual P3 amplitude (394).
The contradictory findings about P3 may be, in parts, due to discrepancy in recording procedures
and differences in experimental criteria across studies (84). More importantly, a sole focus on P3
during oddball paradigm undermines the extensive amount of information on brain dynamic that
can be retrieved form ERPs at millisecond resolution. Particularly, different pathologies in
28
preclinical AD may manifest as abnormalities in specific ERP patterns which have remained
unexplored. Thus, efforts must be focused towards further improvement and characterization of
ERP patterns for establishing ERP as a low-cost, potentially portable, and noninvasive biomarker
for preclinical AD.
1.7 Strategies to improve ERP biomarkers
One strategy to achieve this is to use a behavioral paradigm that depends on brain regions
vulnerable to AD pathology. One such paradigm is trace eyeblink conditioning (TEBC) paradigm
(395). The paradigm consists of a neutral conditioned stimulus (CS) that is paired with an eyeblink-
eliciting stimulation (US) to the eyelid following a stimulus-free interval of 500 ms, and the
conditioned response (CR) is defined as preemptive eyeblink occurring after the onset of the CS
and prior to onset of the US. Indeed, studies have reported that patients with AD perform
significantly worse than age-matched controls in TEBC (396, 397).
The contributing factors to TEBC impairment in AD have been elucidated from studies using
knock-out mice or systemic pharmacological blockade of mAChRs in rabbits and humans, which
showed that cholinergic input mediated through mAChRs is necessary for acquisition in TEBC
(398, 399). Further, a recent study elucidated the role of EC as the key requirement for acquisition
in TEBC, by demonstrating that inhibition of EC with local infusion of γ-aminobutyric acid
(GABA) receptor agonist in rats prior to conditioning prevented them from acquiring a CR (400).
The same study also showed that EC-dependent memory acquisition in TEBC is modulated by
local mAChRs, as infusion of scopolamine into the EC before conditioning slowed acquisition in
rats (400). Consistent with the cholinergic hypothesis of AD, these findings indicate that
impairment of TEBC in AD patients is, in parts, likely due to disruption of BF muscarinic
cholinergic input to the EC and hippocampus (401, 402).
Moreover, a recent study in rats showed that pattern of cortical oscillations elicited by TEBC are
sensitive for accumulation of HP-tau in the entorhinal cortex (403). By using viral vector
approach, a human mutated tau gene was transduced to a subset of cells in the EC of adult rats.
Comparable to the control rats, the HP-tau-expressing rats acquired associative memory in TEBC,
but exhibited an abnormal pattern of local field potentials (LFPs) in the hippocampus and mPFC
29
(403). This abnormality, which was absent in control rats transduced with green fluorescent protein
(GFP) gene, included an attenuation of hippocampal and prefrontal theta power and elevation of
prefrontal beta power. Together, these findings indicted that one of the earliest pathological
features in AD brains, accumulation of HP-tau in the EC, induced abnormal LFP pattern without
memory impairments (403).
In a modified version of TEBC, known as differential trace eyeblink condition (DTEBC), one CS
(CS+) is paired with the US while the other CS remains neutral (CS-). After initial acquisition
phase, CS-US contingency can be switched to study reversal learning. Findings of a study on
rabbits which received bilateral hippocampal lesion before undergoing two-tone DTEBC, showed
that damage to hippocampus greatly impaired reversal learning, but produced a mild effect on
acquisition (404). Severe impairment of reversal learning was also reported by a similar study in
which rabbits with bilateral hippocampal lesion underwent tone-light DTEBC (405). Since the
mPFC receives projections from the hippocampus, it was speculated that damage to the mPFC
may produce similar deficits in reversal learning as observed in rabbits with hippocampal lesions.
This was tested in a study where rabbits received bilateral lesions to the mPFC prior to undergoing
DTEBC (406). Lesions to mPFC caused a significant impairment of reversal learning in rabbits,
but similar to previous findings, performance during acquisition was not significantly affected
(406). Moreover, impairment of reversal learning has also been reported in amnesic patients with
selective hippocampal damage who showed intact learning during the acquisition phase of a visual
discrimination task (407). Together, these findings support a role for hippocampus and mPFC in
reversal learning, and indicate that having additional paradigm to a simple acquisition phase may
enhance the ability to detect subtle sign of memory impairments.
Since TEBC and DTEBC paradigms in animals can be applied to human subjects without any
major modification, these results hold a significant translational potential. However, there remains
the need to investigate how the observed aberrant LFPs translate to ERP measures, which are
known to be comparable between rats and humans (408-410). To this end, the proposed research
will examine the effects of pathological features of early-stage AD on ERPs while rats learn a
modified version of TEBC paradigm.
30
1.8 Hypothesis and Approach
The main hypothesis of this study is that different pathological features of preclinical stage AD
induce distinct abnormal ERP patterns during trace eyeblink conditioning to a degree which is
sufficient for an accurate prediction of the type of brain pathology based on the ERP patterns.
To address this hypothesis, we measured cortical EEG while rats with entorhinal HP-tau over-
expression, cholinergic deficiency, or both conditions, and their associated controls received
differential trace eyeblink conditioning in two phases of acquisition and reversal. In doing so, first
we examined the effect of early pathological features of AD on memory acquisition, stimulus
discrimination, and reversal learning.
Next, we examined ERP patterns in the frontal, parietal, and temporal cortices by looking at peak
amplitude and peak latency of all the visible components, including P1, N1, P2, and P3 potentials.
Subsequently, we submitted all the collected ERP data to a machine learning algorithm and
identified distinct ERP abnormality sensitive for entorhinal HP-tau over-expression, cholinergic
deficiency, and both conditions in conjunction.
31
Chapter 2 Methods
2.1 Subjects
Forty experimentally naïve male Long-Evans rats (Charles River Laboratories, St.-Constant, QC,
Canada), 70 days old upon arrival, were individually housed in transparent plastic cages (45 cm
long x 25 cm wide) in a home colony room and maintained on a reversed 12-hour light – 12-hour
dark cycle (dark from 10:00 to 22:00) with free access to food and water and behavioral testing
occurring during the dark phase, typically between 12-4 pm. This study was approved by
University of Toronto’s Institutional Animal Care Committee
2.2 Design
The study consisted of four groups of animals. Two groups of animals had overexpression of
hyperphosphorylated human tau (HP-tau) protein within the entorhinal cortex and the other two
groups expressed green fluorescent protein (GFP) in the same region as control. Twenty minutes
before conditioning sessions one HP-tau-expressing group (n=10) and one GFP-expressing
group (n=11) received cholinergic receptor blocker, and the other HP-tau-expressing group (n=8)
and GFP-expressing group (n=11) received saline. All animals underwent the same behavioral
protocol described below.
2.3 Surgeries
Surgeries were performed under aseptic conditions using isoflurane anesthesia (4% isoflurane with
1 L/min oxygen for induction; approximately 2% isoflurane with 0.6 L/min for maintenance of
anesthesia; Holocarbon Laboratories, River Edge, NJ, USA). Immediately after induction, rats
received subcutaneous injections of saline (1 ml/100 g body weight) and carprofen (0.5 ml/kg, 50
mg/L) to prevent dehydration and to treat postoperative pain and inflammation. A small amount
of topical analgesic EMLA was applied to the scalp.
32
2.4 Viral vector infusion
To mimic entorhinal tau pathology observed in preclinical stages of AD (44), a recombinant adeno-
associated viral serotype 9 (Figure 4; rAAV9) vector with mutated human tau gene was bi-laterally
micro-infused in the entorhinal cortex of adult rats to over-express mutated human tau protein. The
form of tau transgene contained the P301L mutation associated with frontotemporal dementia, and
four microtubule-binding domains (exons 2−/3−/10+). The cytomegalovirus (CMV)
enhancer/chicken β-actin promoter combination and the woodchuck hepatitis post transcriptional
regulatory element were used to drive gene expression. The viral vectors were provided by Dr.
Klein who validated the vector in rodents (411). The vector was bilaterally infused using a micro
infusion pump (Harvard Apparatus, Holliston, MA) in four anteroposterior (AP), mediolateral
(ML), and dorsoventral (DV) coordinates (Figure 5; Table 1; 3 μL/site at 0.375 µL/min). The
coordinates were based on our previous study to maximally transduct cells in the entorhinal region
(379). For control, rAAV9-GFP vector was infused into the same coordinates in another set of
animals. Prior to infusion, the titer of vector (rAAV9-tau-P301L or rAAV9-GFP) was prepared to
be ~3.7×1013 vector genome particles/ml. Following infusion, the incision was closed using silk
sutures. The viral vector was incubate for six weeks before initiation of behavioral testing (379).
2.5 EMG-EEG electrode Implantation
Two weeks after recovery from viral vector surgeries, animals were subjected to implantation of
electromyogram (EMG) and electroencephalogram (EEG) electrodes to record muscle and brain
activities respectively during the behavioral paradigm. Following the incision, each rat was
implanted with 3 stainless steel screw electrodes on the skull surface above right frontal (3.2 mm
AP, 0.7 mm ML), parietal (-3.5 mm AP, 2.0 mm ML), and temporal (-6.2 mm AP, 7.0 mm ML)
cortices, and a reference screw was placed on the skull above the right cerebellar cortex (Figure 6
A). Five additional screws were added that served as anchors for the dental acrylic. Next, four
Teflon-coated stainless steel wire (outer diameter = 0.0055”; A-M Systems, Carlsberg, WA) was
inserted in orbicularis oculi muscle of the left eyelid to serve as an EMG electrode pair and a
stimulus electrode pair (Figure 6 B). The recording and reference electrode wires were then
connected to a 9-pin ABS plug (Ginder Scientific, Nepean, Canada) which was fixed to the skull
33
with dental acrylic. Following the implantation of electrodes, the incision was closed up to the
head cap by silk sutures and the EMG electrode wire was uncoated and trimmed at the level of the
eyelid.
2.6 Drug treatment
Scopolamine hydrobromide (Sigma-Aldrich, Oakville, Canada) was dissolved in 0.9% saline. Half
of rats in HP-tau group (n= 8) and GFP group (n=11) received a subcutaneous injection of (0.05
mg/kg body weight) of scopolamine hydrobromide, while the remaining animals in HP-tau group
(n=10) and GFP group (n=11) received subcutaneous 0.9% saline injection 20 minutes before the
conditioning started. The drug treatment began on the second adaptation session and continued
until the last session. To recapitulate the level of cholinergic deficiency in preclinical,
asymptomatic AD stage, the dose of scopolamine was chosen to thwart but not completely block
memory function or neural activity at the level of the brainstem (398, 412).
2.7 Apparatus and Behavioral paradigm
Prior to behavioral testing animals were assigned into four groups of HP-tau and GFP with saline
treatment (tau-sal; GFP-sal) and HP-tau and GFP with scopolamine treatment (tau-scop; GFP-
scop). The groups underwent a variant of classical conditioning paradigm, known as differential
trace eyeblink conditioning (Figure 7; DTEBC). In this paradigm a neutral conditioned stimulus
(CS) was paired with an aversive, eyeblink-eliciting stimulus (US) with an inter-stimulus interval
of 500 millisecond (ms). This paradigm included an auditory stimulus (ACS, 100-ms, 2.5-kHz,
85-dB pure tone) and a visual stimulus (VCS, 100-ms, 50-Hz, LED light). LED was located at the
font of the chamber and a speaker was mounted on the box ceiling directly above the cylinder. The
US was a 100 ms periorbital shock (100 Hz square pulses) delivered to the eyelid with a stimulus
isolator (ISO-Flex, A.M.P.I., Jerusalem, Israel). The shock level was initiated at 0.3 mA and was
adjusted individually and daily for each rat to induce an optimal unconditioned response (UR; an
eyeblink/ head-turn response) which was monitored through web cameras mounted inside the
chambers.
34
During habituation (Days 1-2), animals were placed into a transparent, cylindrical, plastic
container and put into a sound and light-attenuating chamber for 50 minutes without any stimulus
presentation to adapt to the environment, while baseline EMG was recorded. On the second day,
animals received their respective drug treatments to observe whether there would be an effect on
baseline eyelid activity. During the acquisition phase (Days 3-12), rats received 50 pairings of one
CS (CS1+, either the auditory or visual counterbalanced across rats) and the US in addition to 50
presentations of the other CS alone (CS2-). After rats formed the association between the CS1+
and US, reversal sessions began (Days 13-19), in which the CS1 was presented alone (CS1-) while
the CS2 was reinforced with the US (CS2+). EEG and EMG were recorded during all session.
Eyelid EMG was used to detect conditioned responses (CR, defined as eyeblink responses elicited
during a 200 ms window immediately before US onset). Rats’ memory performance was
quantified with the ratio of trials with CR expression to the total number of trials in a session (CR
%).
2.8 Behavioral testing coupled with EEG recording
During conditioning, EMG from the left upper orbicularis oculi and EEG from the frontal, parietal,
and temporal epidural electrodes were continuously recorded. The signals were amplified, band-
pass filtered (300–3000 Hz for EMG; 0.1–400 Hz for LFP), and acquired at 6,102 Hz (EMG) or
2,034 Hz (EEG) with a RZ5 recording system (Tucker-Davis Technology, Alachua, FL). The
OpenEx software (Tucker-Davis Technology, Alachua, FL) was used to monitor the trial events
and insert the time stamps and corresponding comments (CS1, CS2, US, trial number) to the EEG
and EMG data. Arduino Mega 2560 was to initiate the paradigm and administer trial events (CS
and US presentation).
2.9 EMG Analysis
All of the analyses were performed using custom codes written in Matlab (Mathworks, Natick,
MA, USA). The method of analysis of the EMG data was the same as in our previous study (376,
413). The instantaneous EMG amplitude was defined as the difference between the minimum and
maximum EMG signal in a 1-ms window. Noise was removed from the EMG amplitude by
35
subtracting the across-trial mean, plus one SD, of the amplitude during a 300-ms period before the
CS onset. All trials in which the pre-CS EMG amplitude exceeded the noise level by an additional
30% of this noise level were excluded from further analyses. A trial was defined as containing a
CR if the EMG amplitude during a 200-ms window prior to the US (CR amplitude) exceeded five
times the pre-CS amplitude (in the case that the pre-CS amplitude was zero, CR amplitude had to
be 10% greater than the noise level). The CR percent was defined as the ratio of the number of
trials containing the CR to the total number of valid trials in each session.
2.10 EEG analysis
EEG data from each session were then processed by using custom codes written in Matlab
(MathWorks, Novi, USA). EEG data from individual rats were first checked for the contamination
of power line noise by examining the EEG power spectrum. EEG data was excluded from future
analysis if the size of the peak amplitude at the 60 Hz frequency range was more than half the size
of the highest peak amplitude between the 5-10 Hz frequency band. To calculate event-related
potentials (ERPs), we extracted EEG activity during 6 second around the onset of CS and averaged
the EEG data across all trials in one session. Then, ERPs were averaged across the last three
acquisition and reversal sessions separately for two CS types (tone and light) and three brain
regions (frontal, parietal, and temporal) in each rat. Next, from the averaged ERPs, we then
calculated the amplitude and latency (2 features) of P1, N1, P2, and P3 components. These
components were detected as the peak of positive (P1-3) or negative (N1) deflection occurred
during a set time window after CS onset (Figure 8). For auditory ERPs, the time window for P1
was 35-80 ms, for P2 was 120-240 ms, for P3 was 240-360 ms, and for N1 was 80-120 ms. For
visual ERPs, the time window for P1 was 90-130 ms, for P2 was 180-260 ms, for P3 was 260-360
ms, and for N1 130-180 ms. These time windows were set based on the previously reported ERP
patterns induced by auditory (381, 382) and visual stimuli (380) in humans and rats. The detected
amplitude and latency were averaged across rats in each of four groups.
36
2.11 Support Vector Machine
To examine whether each type of pathology induces a distinct pattern of ERP abnormalities, we
applied a Support Vector Machine (SVM) classifier on ERP features by using LIBSVM package
(414). The SVM is a supervised machine learning algorithm which are used for data classification
(415-419). With given set of data containing observations whose categories are known, SVM’s
training algorithm builds a model that identifies to which set of categories a new unlabeled
observation belongs. The SVM model creates a representation of the labeled observations as points
in space, in a way that observations belonging to different categories are separated by a decision
surface with maximum margin. The location of the decision surface is determined by the data
points (support vectors) that lie closest to it. The margin between the decision surface and the
closest data points from each category is maximized to reduce the number of misclassifications.
New sets of unlabeled observations are then mapped into the same space and are predicted to
belong to a category based on the side of the decision surface they fall on.
The classification procedure involved four steps. From each ERP data, we extracted the amplitude
and latency of P1, P2, P3, and N1 components (defined the previous section). We repeated this
procedure for three brain regions (frontal, parietal, and temporal), for two CS modality (tone,
light), and for two association type (CS-, CS+). This results in total 96 features for each rat (Figure
9). The extracted ERP features were first transformed to the format of an SVM package. All rats
within a single group were organized in rows and were assigned the same class labels (1 to 4 for
GFP-saline, tau-saline, GFP-scopolamine, and tau-scopolamine respectively). The corresponding
features for each rat were placed on the same row level and organized in columns 1 to 96 (Figure
10). Subsequently, the data was linearly scaled to the range [0, 1] to prevent features in greater
numeric ranges (i.e. amplitudes) dominate those in smaller numeric ranges (i.e. latencies). As such,
the values of each feature across all rats were divided by the maximum value of that feature. Next,
a radial basis function (RBF) kernel was used to train the classifier and classify the data (389).
RBF kernel was chosen because of its ability to nonlinearly map samples into a higher dimensional
space if class labels and features follow a nonlinear relation, and has been used in other studies for
classification of neural data (393, 394).
The performance of a given SVM with an RBF kernel depends on the two parameters of the RBF
kernel C and γ, however it is not known beforehand which C and γ pair optimize SVM’s
37
classification of the data. To find the best pair of C and γ, a grid-search with a five-fold cross-
validation rate was executed on a sequence of (C, γ) pair values that grew exponentially (i.e. C =
2-5, 2-3,… 215, γ =2-15, 2-13, …23). The parameter (C, γ) pair that produced the highest cross-
validation accuracy was picked. The classifier was then trained using the best parameter (C, γ) pair
with the data from all rats except one, and was tested on the remaining rat. The classification
accuracy was defined as the probability of correctly classifying a test rat, and computed as the
percentage of correctly classified instances. Lastly, the 95% confidence intervals for SVM
classification accuracy was calculated using the binofit function in Matlab, and served as a decision
boundary for testing whether the observed classification accuracy was significantly better than
chance.
2.12 Statistical analysis
The effects of scopolamine, entorhinal HP-tau over-expression, and both conditions together, on
the expression of a conditioned blinking response were assessed using a three-way repeated
measures analysis of variance (three-way RM-ANOVA), with groups (i.e. GPF-saline, GFP-
scopolamine, tau-saline, and tau-scopolamine) as the between-subjects factor and sessions (i.e.
acquisition or reversal) and CS types (CS+, CS-) as the within-subjects factor. Data are presented
as the mean ± the standard error of the mean (SEM). Differences were considered statistically
significant if p <0.05. Next, the effects of scopolamine, entorhinal HP-tau overexpression, and
both conditions together, were examined on amplitude and the latency of brain’s ERPs. Due to the
known difference in ERPs to auditory and visual stimuli (420, 421), ERPs to the auditory (or
visual) CS during the acquisition phase were compared against those during the reversal phase.
The ERPs for the last three conditioning sessions of acquisition and reversal were first averaged
in each rat, and subsequently averaged across all the rats in each treatment group. The amplitude
and latency values of each ERP component were compared between treatment groups by three-
way RM-ANOVA with virus (i.e. GFP, tau) and drug (i.e. saline, scopolamine) as the between-
subjects factors and CS type (CS+, CS-) as the within-subjects factor. Data are presented as the
mean ± the standard error of the mean (SEM Bonferroni corrected p-values were used for multiple
comparisons and differences were considered statistically significant if p <0.00625. All analyses
were performed using IBM SPSS Statistics software (SPSS, San Jose, USA).
38
2.13 Immunohistochemistry
Once behavioral testing was finished, viral vector expression and spread were verified by
immunohistochemistry. Each rat was given a lethal intraperitoneal injection of sodium
pentobarbital (102 mg/kg; Euthansol, Merck, Kirkland, QC, Canada) approximately after 10 weeks
of incubation and perfused intercardially with approximately 150 mL of 0.9% saline and 150 mL
of 4% paraformaldehyde. The brains were then extracted, post-fixed for 2 hours in 4%
paraformaldehyde solution at 4 oC, and transferred to a solution of 30% sucrose in PBS (0.1 M;
pH 7.4) for 1 week. Subsequently, brains were sliced to sections 40 um thick using a cryostat
equipped with a freezing-sliding microtome (Leica Microsystems, Canada). The presence of HP-
tau associated with pretangle material and neurofibrillary changes in AD was assessed
immunohistochemically using monoclonal antibody, AT8 (mouse; 1:1000; #MN1020,
ThermoFisher Scientific) which detected human tau protein with phosphorylated serine 202 and
threonine 205 amino acid residues (pSer202/Thr205) (422). Cell nucleus was visualized with the
DAPI stain (EMD Millipore).
Images were acquired using an upright fluorescent microscope (Leica Microsystems, Canada) and
Volocity image analysis software (PerkinElmer, Waltham, USA), and were mapped with the
stereotaxic atlas of the rat brain (423). The behavioral and EEG data from a rat were included in
the subsequent analyses if imaging results showed 1) an accurate placement of the infusion cannula
within the entorhinal cortex in all the injection sites, and 2) HP-tau was detected in minimum 6
out of 8 of total correctly-placed injection sites. The volume of entorhinal cortex inflicted with
HP-tau was obtained by estimating the volume of expression using the mediolateral, dorsoventral,
and anteroposterior lengths of spread in a series of sections collected every 80 m.
39
Coordinates from bregma (mm)
AP ML DV
Site 1 6.0 6.65 -8.8 Site 2 6.8 6.4 -8.0 Site 3 7.6 5.6 -7.2 Site 4 8.5 5.4 -6.8
Table 1. Coordinates of four injection sites where viral vector containing either human mutated tau or GFP is infused into the entorhinal cortex (Anteroposterior, AP, Mediolateral, ML; Dorsoventral, DV).
Figure 4. A recombinant adeno-associated viral serotype 9 (rAAV9) with the
cytomegalovirus (CMV) enhancer and chicken β-actin promoter were used to drive the
expression of human tau with P301 mutation in the model rats, or green fluorescent protein
(GFP) in the control rats. The vectors were obtained from Dr. R. Klein at the Louisiana
State University, Baton Rouge, United States.
Figure 4. A cartoon representation of rAAV9 vector.
CMV
Enhancer Chicken β-actin promoter
Human tau P301L
gene
Translation starts Translation stops
Model rats: rAAV2/9 Tau
Control rats: rAAV2/9 GFP
CMV
Enhancer Chicken β-actin promoter
Green fluorescent protein
gene
Translation starts Translation stops
Figure 5. Bi-lateral infusion of the viral
vector into the entorhinal cortex.
Table 1. Coordinates of the viral
vector infusion.
Figure 5. A depiction of a histological
section from a single rat illustrating the
first of the four locations of the micro-
injection within the entorhinal cortex.
40
EEG Frontal
Parietal
Temporal
EMG &
Shock
US
Time
Figure 6. A) stainless steel screw electrodes were
implanted on rat’s skull surface and penetrated the
dura mater to record the cortical activity over the
frontal (3.2 mm anterior and 0.7 mm lateral from
bregma; yellow), parietal (-3.5 mm AP, 2.0 mm ML;
red), and temporal (-6.2 mm AP, 7.0 mm ML; blue)
cortices as shown in the sample extracted brain. B)
A depiction of EEG and EGM acquisition screen.
Four Teflon-coated stainless steel wire (outer
diameter = 0.0055”) were inserted in orbicularis
oculi muscle of the left eyelid to serve as an EMG
electrode pair and a stimulus electrode pair.
Figure 6. EEG and EMG electrode placements.
B
(Adapted with few modifications, Madronal et al., 2009.)
Anchor screw
indentation
41
Figure 7. A depiction of differential trace paradigm. In this paradigm two stimuli (tone and light)
are presented at random trials with equal probability. In phase 1 (acquisition) rats received 50
pairings of the neutral conditioned stimulus 1 (CS1) with an aversive, eyeblink-eliciting stimulus
(US) with an inter-stimulus interval of 500 ms, and 50 presentations of the other CS alone (CS2-).
In phase 2 (reversal) CS1 was presented alone (CS1-) while the CS2 was reinforced with the US
(CS2+).
Figure 7. Differential trace eyeblink paradigm.
42
Figure 8. Time windows for capturing ERP peaks after CS onset.
Figure 8. The time window after CS onset during which peaks for different auditory
and visual ERP components are captured. The time windows for auditory ERP
components include, 35-80 ms for P1, 80-120 ms for N1, 120-240 ms for P2, 240-360
ms for P3 after the CS onset. For visual ERP components the time windows are 90-
130 ms for P1, 130-180 ms for N1, 180-260 ms for P2, and 260-360 ms for P3
component after the onset of the CS.
Visual ERPs
+
Am
plit
ud
e (A
U)
-
Auditory ERPs
+
Am
plit
ud
e (A
U)
-
43
Figure 9. ERP feature extraction and data set combinations for SVM
Figure 9. A) A total of 96 ERP features were extracted from
the combination of 2 stimulus modalities, 2 stimulus
association with the US, 2 features of ERP peaks, 3 brain
regions, and 4 ERP components from each animal. B) The
extracted features were used as the input data for support-
vector machine (SVM) classifier in 14 different combinations
to examine which feature allows SVM to classify different
treatment groups with the highest level of accuracy.
44
Figure 10. Summary of steps in SVM classification.
Figure 10. A) Extracted ERP features were transformed into the format of an SVM package before
classification of the data. All rats within a single group were organized in rows and were assigned the same
class labels (1 to 4 for GFP-saline, tau-saline, GFP-scopolamine, and tau-scopolamine respectively). The
corresponding features for each rat were placed on the same row level and organized in columns 1 to 96. B)
The data for each feature was then scaled from 0-1 across all rats. C) Using a five-fold cross validation, 143
pairs of C & γ were tested with the data set to obtain the pair with highest cross-validation accuracy D) Using
leave-one-out cross validation (LOOCV) one rat is left out while the remaining data set was trained with the
best C & γ pair and the model was tested with the sample which was left out. This procedure was repeated until
the model was tested once with all the samples (Round 1-40; rat 1-40). The classification accuracy was defined
as the probability of correctly classifying a test rat, and computed as the percentage of correctly classified
instances.
A) Data set in SVM format B) Data set scaled 0-1
Fold 1 Fold 2 Fold 3 Fold 4 Fold 5
143 (C & γ) pairs ×
D) Leave-one-out cross validation with the best (C & γ) pair
Round 2: Take rat 2 for test,
Train with rats 1, 3-40
Round 1: Take rat 1 for test,
Train with rats 2-40
Round 3-40: Take next rat for test,
Train on rest plus rats 1&2
Rat number Rat 1
Rat 40
45
Chapter 3 Results
3.1 Entorhinal hyperphosphorylated tau over-expression
Histological analyses was conducted at the end of behavioral testing and approximately 10 weeks
after incubation of the viral vector with the antibody AT8 targeted to human tau phosphorylated
serine residues 202 and 205. From 24 rats that received the rAAV-tau infusion, 4 rat did not show
an expression of HP-tau protein and 4 rats had misplaced infusion cannula resulting in expression
of HP-tau in ventral hippocampus or perirhinal cortex. Also, 2 of the 24 control rats did not show
any GFP expression. The data from these animals were removed from subsequent analyses.
Examination of HP-tau expression in the entorhinal cortex (EC) revealed that a sizable fraction of
cells in in the EC expressed HP-tau protein (Figure 11 A). Higher magnification images further
revealed the presence of HP-tau inside the dendrites, soma, and axons of entorhinal cells (Figure
11 B).
To estimate the spread of cells expressing HP-tau, we examined the location of tau-expressing
cells in a series of 40 um sections from a single rat (Figure 11 C). HP-tau was observed most
intensely in areas adjacent to injection sites, but it also spread into areas between the injection sites
(Figure 11 D).
In parallel, when examining the cortical areas dorsal to the entorhinal cortex (Figure 11 E), we did
not observe any expression of HP-tau in any of these cortical regions at or between the injection
sites, except for some minor expression along the injection tract and a small mediolateral (ML)
spread of approximately 0.13 ± 0.02 mm into the perirhinal cortex for the last two injection
coordinates (AP -7.6 and -8.5).
Based on the location of the entorhinal HP-tau-expressing cells (Figure 11 D), we measured the
ML spread of the HP-tau to be approximately 0.69 ± 0.11 mm at and between the injection sites
and the anteroposterior (AP) spread to be 3.5 mm. Subsequently, by using ML and AP spread
values we estimated the volume of expressed HP-tau to be 47.1% of total entorhinal volume.
46
Hyperphosphorylated tau is known to spread transynaptically (424, 425). To test whether HP-tau
spread to regions that receive entorhinal axonal projections, we examined the expression of HP-
tau in two of entorhinal efferent regions, the dorsal hippocampus (424) and prelimbic region of
medial prefrontal cortex (424, 426). Our examination did not reveal any expression of HP-tau in
the dorsal hippocampus or the prelimbic region of the prefrontal cortex (Figure 11 F)
Together, these results suggest that we successfully induced expression of HP-tau in rat brain.
Because tau aggregation is limited to the entorhinal cortex, it corresponds to transentorhinal stage
in which the pathology is mostly restricted in the entorhinal and perirhinal regions (1).
47
A) Sizeable fraction of cells in the entorhinal
cortex expressed hyperphosphorylated tau
Figure 11. Expression of hyperphosphorylated tau in the entorhinal cortex.
Figure 11. A) Examination of HP-tau expression in the entorhinal cortex (EC) revealed that a sizable fraction of
cells in in the EC expressed HP-tau protein (red). B) Magnified images of AT8-immunolabelled EC sections
(red) of tau-expressing rats showed extensive expression of hyperphosphorylated tau within the cell body (20X,
40X) as well as dendrites and axons of entorhinal cells (40X). Cell nucleus was visualized by DAPI stain (blue).
B) Higher magnification images of entorhinal cells expressing hyperphosphorylated tau
40 µm 20 µm
320µm
48
Position relative to Bregma (mm)
D) Expression and spread of mutated
tau protein at the entorhinal cortex C) Coronal sections of areas before and at the injection sites
Figure 11. C) Coronal sections from the rat brain atlas corresponding to sections of entorhinal
cortex (red) at the site of bilateral viral infusion (red square) as well as the area in between (pink
square). D) AT8-immunolabelled images of entorhinal sections from a single tau-expressing rat
exhibit a consistent expression of tau throughout the dorsolateral EC, as well as expressions in the
medial and central EC from -8.00 to -8.5 anteroposterior coordinate from bregma.
-5.50
-6.00
-6.40
-6.80
-7.20
-7.60
-8.00
-8.50
900 µm
49
Figure 11. E) Images were taken from cortical areas dorsal to the entorhinal cortex at and
between the injection sites, including the ectorhinal cortex, temporal association cortex,
and the primary and secondary auditory cortices to examine the spread of HP-tau into the
neighboring regions. No expression of mutated tau (red) was observed in these cortical
regions at or between the injection sites, except for some minor expression along the
injection tract and a small mediolateral ML spread of approximately 0.13 ± 0.02 mm into
the perirhinal cortex for the last two injection coordinates (AP -7.6 and -8.5).
Position relative
to Bregma (mm)
E) Examination of mutated
tau expression in cortical areas
dorsal to the entorhinal cortex
900
µm
50
Figure 11. F) Expression of viral vector carrying GFP (green) or mutated tau (red). Immunolabeling of brain
sections from the prelimbic (PrL) region, dorsal hippocampus (dHPC), and entorhinal cortex (EC) of GPF and
hyperphosphorylated tau-expressing rats with AT8 antibody targeting phosphorylated tau protein revealed
expression of GFP and tau only at the site of viral infusion within the entorhinal cortex, and not in PrL or dHPC.
F) Expression of the viral vector confined to the entorhinal cortex
51
3.2 Behavior
As expected, there was no significant difference between the groups during the two habituation
sessions (three-way repeated measures ANOVA, Sessions × Virus × Drug interaction, F (1, 36)
=0.075, p = 0.786; Sessions × Virus, F (1, 36) =0.160, p = 0.692; Sessions × Drug, F (1, 36) =0.119, p
= 0.732).
3.2.1 Acquisition
The performance of animals in all groups were comparable through the 10 days of acquisition in
DTEBC (Figure 12 A-E; three-way repeated measures ANOVA, CStype × Session × Group
interactions, F (27, 324) = 0.704, p= 0.863; CStype × Group, F (3, 36) = 2.273, p = 0.097; Session ×
Group, F (27, 324) = 1.070, p = 0.375; Group, F (3, 36) = 0.219, p= 0.882). Through the 10 days of
acquisition, all groups significantly increased the frequency of their expression of the conditioned
response (CR %) to the reinforced conditioned stimulus (CS1+), but not in unreinforced CS (CS2-
) (CStype × Session, F (9, 324) = 22.153, p < 0.001; follow-up one-way RM-ANOVA for GFP-saline,
Session, F(9,90) = 8.162, p < 0.001; one-way RM-ANOVA for GFP-scopolamine, Session, F (9,117)
= 8.788, p < 0.001; one-way RM-ANOVA for tau-sal, Session, F (4.109, 41.087) = 8.535, p < 0.001;
one-way RM-ANOVA for tau-scopolamine, Session, F (3.422, 30.801) = 7.616, p < 0.001). Further, all
groups learned the task at a similar rate. There was no significant difference in number of sessions
required for rats in different groups to show a difference of 50% CR between the CS1+ and CS2-
(Virus × Drug, F (1, 36) =0.188, p=0.667; Virus, F (1, 36) =1.282, p=0.265; Drug, F (1, 36) =1.994, p =
0.166). Overall, rats with entorhinal HP-tau over-expression, scopolamine treatment, or both in
conjunction, were able to acquire an associative memory with similar rates to the control rats.
52
3.2.2 Reversal
The performance of animals in all groups were comparable through the 7 days of reversal in
DTEBC (Figure 13 A-E; three-way repeated measures ANOVA, CStype × Session × Group
interactions, F (18, 216) = 0.921, p= 0.554; CStype × Group, F (3, 36) = 10.46, p= 0.384; Session ×
Group, F (18, 216) = 0.852, p= 0.637; Group, F (3, 36) = 0.096, p = 0.962). Through the 7 days of
reversal phase, all groups were able to significantly increase their CR % to the newly-reinforced
conditioned stimulus (CS2+), but not in unreinforced CS (CS1-) (Figure 13 A-E; CStype ×
Session, F (6, 216) = 25.66, p < 0.001; one-way RM-ANOVA for GFP-saline, Session, F (6, 33.347) =
8.939, p < 0.001; one-way RM-ANOVA for GFP-scopolamine, Session, F (3, 43.352) = 10.633, p <
0.001; one-way RM-ANOVA for tau-sal, Session, F (3.294, 32.936) = 5.922, p =0.002; one-way RM-
ANOVA for tau-scopolamine, Session, F (1.921, 17.289) = 6.218, p =0.010). Further, animals in all
groups learned the task at a similar rate. There was no significant difference in number of sessions
required for rats in different groups to show a difference of 50% CR between the CS2+ and CS2-
(Virus × Drug, F (1, 36) =0.040, p=0.843; Virus, F (1, 36) =0.303, p = 0.586; Drug, F (1, 36) =1.692, p =
0.202). Together, the results of rats’ performance during acquisition and reversal phase indicate
that associative learning and reversal learning are not impaired by entorhinal HP-tau over-
expression, scopolamine treatment, or both conditions together.
53
Figure 12. DTEBC Acquisition (10 Days).
Figure 12. Differential trace eyeblink conditioning in GFP-saline (A, red), GFP-scopolamine (B,
pink), tau-saline (C, dark blue), and tau-scopolamine (D, light blue) groups. All groups underwent
2 days of habituation (H1-2) and 10 days of acquisition (A1-10). All groups were able to acquire
a conditioned response (CR) to the reinforced stimulus (CS1-US; solid line) with comparable rates,
and did not develop a CR to the non-reinforced stimulus (CS2-alone; dashed line).
A) GFP-Saline B) GFP-Scopolamine
C) Tau-Saline
E) All groups
Sessions
CR
%
D) Tau-
Scopolamine
0 H1 H2 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
D) Tau-Scopolamine
CS1+ CS2-
CS1+ CS2-
CS1+ CS2-
CS1+ CS2-
CR
%
CR
%
54
Sessions
A) GFP-Saline B) GFP-Scopolamine
C) Tau-Saline D) Tau-Scopolamine
E) All groups
CR
%
Figure 13. DTEBC Reversal (7 Days).
Figure 13. Differential trace eyeblink conditioning in GFP-saline (A, red), GFP-scopolamine (B, pink), tau-saline (C,
dark blue), and tau-scopolamine (D, light blue) groups. All groups underwent 7 days of reversal (R1-7), where the
contingency of the stimuli were switched (CS2-US, CS1-alone). All groups acquired conditioned response (CR) to
the reinforced stimulus (CS2-US; solid line) through the 7 days of reversal phase. GFP-saline (A) and tau-saline (B)
differentiated between the CS2-US and CA1-alone by day 4 of the reversal phase, while GFP-scopolamine (C) and
tau-scopolamine (D) became able to differentiate between the stimuli by day 5 and 7 of the reversal respectively.
CS2+ CS1-
CS2+ CS1-
CS2+ CS1-
CS2+ CS1-
CR
%
CR
%
55
3.3 Event-related potentials
Examination of EEG power spectrum of each individual rat recorded in the last three days of
acquisition or reversal phase did not reveal any power line noise artifact. As such, all the EEG data
for each individual rat recorded during these sessions were used to calculate the ERPs.
3.3.1 Auditory stimulus-evoked ERPs in the frontal cortex
In GFP-saline group, the frontal auditory ERPs displayed a pronounced P1 and N1 deflection, and
a smaller P2 deflection (Figure 14 A). Compared to the GFP-saline group, the amplitude of P1
was larger in GFP-scopolamine and tau-scopolamine group, but not in tau-saline group (Figure 15
A; three-way RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =27.532, p < 0.001; follow-
up independent sample t-test, GFP-saline vs. tau-saline, t= 2.370, p= 0.030; GFP-saline vs. tau-
scopolamine, t= 3.140, p=0.005; GFP-saline vs. GFP-scopolamine, t= -0.312, p= 0.759).
Moreover, the P1 amplitude in tau-scopolamine was significantly larger than that in the tau-saline
group (t= -5.2, p< 0.001). The latency of P1 was comparable across the groups (Figure 16 A; all
p > 0.05).
Compared to the GFP-saline group, the amplitude of N1 was larger in tau-scopolamine and GFP-
scopolamine group, but not in tau-saline group (Figure 14 D; three-way RM-ANOVA, CStype ×
Virus × Drug interaction, F (1, 36) =8.615, p=0.006; follow-up independent sample t-test, GFP-saline
vs. tau-saline, t= -2.374, p=0.030; GFP-saline vs. tau-scopolamine, t=1.896, p = 0.759; GFP-saline
vs. GFP-scopolamine, t=-0.071, p=0.944) Moreover, the N1 amplitude in tau-scopolamine was
significantly larger than that in the tau-saline group (t = -5.2, p < 0.001). The latency of N1 was
comparable across the groups (Figure 16 D; all p > 0.05).
Further, compared to the GFP-saline group, the amplitude of P2 was smaller in tau-saline group,
but not in GFP-scopolamine group and tau-scopolamine group (Figure 14 G; three-way RM-
ANOVA, CStype × Virus × Drug interaction, F (1, 36) =6.614, p=0.014; Virus × Drug interaction,
F (1, 36) =11.698, p=0.001; follow-up independent sample t-test, GFP-saline vs. tau-saline, t= 6.923,
p < 0.001; GFP-saline vs. tau-scopolamine, t=-1.393, p=0.741; GFP-saline vs. GFP-scopolamine,
t= 0.042, p=0.967) Moreover, the P2 amplitude in tau-scopolamine was significantly larger than
56
that in the tau-saline group (t (13.78) = -6.254, p < 0.001). The latency of P2 was comparable across
the groups (Figure 16 G; all P > 0.05).
Overall, the pattern suggests that while entorhinal HP-tau over-expression results in an attenuation
of the auditory ERP amplitudes in the frontal cortex, scopolamine alone seems to elevate the level
of ERP amplitudes, and even more so in conjunction with HP-tau.
3.3.2 Auditory stimulus-evoked ERPs in the parietal cortex
In GFP-saline group, the parietal auditory ERPs displayed a pronounced P1 and N1 deflection,
and a smaller P2 and P3 deflection (Figure 14 B). Compared to the GFP-saline group, the
amplitude of P1 was smaller in GFP-scopolamine, tau-saline, and tau-scopolamine groups
respectively. (Figure 15 B; three-way RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36)
=0.112, p=0.739; main effect of scopolamine, F (1, 36) =11.344, p=0.002; main effect of virus, F (1,
36) =36.789, p < 0.001). The latency of P1 was comparable across the groups (Figure 16 B; all p >
0.05).
Compared to the GFP-saline group, the amplitude of N1 was smaller in GFP-scopolamine, tau-
saline, and tau-scopolamine groups respectively (Figure 15 E; three-way RM-ANOVA, CStype ×
Virus × Drug interaction, F (1, 36) =1.504, p = 0.228; main effect of scopolamine, F (1, 36) =11.268,
p = 0.002; main effect of virus, F (1, 36) =66.726, p < 0.001). Further, compared to GFP-saline
group, the latency of N1 component was shorter in GFP-scopolamine and tau-scopolamine groups
(Figure 16; E; three-way RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =0.003, p =
0.514; main effect of scopolamine, F (1, 36) =17.072, p < 0.001).
Additionally, compared to the GFP-saline group, the amplitude of P2 was smaller in tau-saline
group, but not in GFP-scopolamine group and tau-scopolamine group (Figure 15 H; three-way
RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =18.1069, p<0.001; follow-up
independent sample t-test, GFP-saline vs. tau-saline, t= -2.783, p= 0.022; GFP-saline vs. tau-
scopolamine, t= 0.307, p = 0.762; GFP-saline vs. GFP-scopolamine, t= -2.632, p=0.016)
Moreover, the P2 amplitude in tau-scopolamine was not significantly larger than that in the tau-
saline group (t (13.78) = -2.784, p = 0.013). The latency of P2 was comparable across the groups
(Figure 16 H; all P > 0.05).
57
Compared to the GFP-saline group, the amplitude of P3 component was smaller in all the
remaining groups (Figure 15 K; three-way RM ANOVA, CStype × Virus × Drug interaction, F (1,
36) =1.888, p = 0.178; Virus × Drug interaction, F (1, 36) =13.068, p < 0.001; follow-up independent
sample t-test, GFP-saline vs. tau-saline t (15.193) =5.37, p <0.001; GFP-saline vs. tau-scopolamine,
t= -3.597, p= 0.002; GFP-saline vs. GFP-scopolamine, t (20) =3.649, p = 0.001; tau-saline vs. tau-
scopolamine t (20) =-1.437, p = 0.177). The latency of P3 was comparable across the groups (Figure
16 K; all p > 0.05).
Overall, the pattern suggests that entorhinal HP-tau over-expression as well as scopolamine
attenuates of all the parietal auditory ERP amplitudes. The effect on ERP amplitudes is further
potentiated when both conditions are present in conjunction.
3.3.3 Auditory stimulus-evoked ERPs in the temporal cortex
In GFP-saline group, auditory temporal ERP displayed a large P1 and N1 deflection, and small P2
and P3 deflection (Figure 14 C). Compared to GPF-saline group, the amplitude of P1 component
was smaller in GFP-scopolamine, tau-saline, and tau-scopolamine respectively (Figure 15 C;
three-way RM ANOVA, CStype × Virus × Drug interaction, F (1, 36) =1.60, p = 0.214; Virus ×
Drug interaction, F (1, 36) =9.288, p < 0.004; follow-up independent sample t-test, GFP-saline vs.
tau-saline t =6.610, p < 0.001; GFP-saline vs. tau-scopolamine, t= -3.597, p= 0.002; GFP-saline
vs. GFP-scopolamine, t =-8.146, p < 0.001) Further, the P1 amplitude in tau-saline group was
significantly larger than that in the tau-scopolamine group (t =4.883, p < 0.001). The latency of
P1 was comparable across the groups (Figure 16 C; all p > 0.05).
Compared to GFP-saline group, the amplitude of N1 component was comparable in tau-
scopolamine groups, but smaller in GFP-scopolamine and tau-saline groups (Figure 15 F; three-
way RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =1.764, p = 0.192; Virus × Drug
interaction, F (1, 36) =132.336, p < 0.001; follow-up independent sample t-test, GFP-saline vs. tau-
saline t =15.36, p < 0.001; GFP-saline vs. tau-scopolamine, t= -1.961, p=0.077; GFP-saline vs.
GFP-scopolamine, t =12.93, p < 0.001) Further, the N1 amplitude in tau-scopolamine group was
significantly larger than that in the tau-saline group (t =8.299, p < 0.001). The latency of N1 was
comparable across the groups (Figure 16 F; all p > 0.05).
58
Compared to GFP-saline group, the largest auditory temporal P2 amplitude was observed in tau-
scopolamine group, followed by GFP-scopolamine and tau-saline groups (Figure 15 I; three-way
RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =0.522, p = 0.474; main effect of
scopolamine, F (1, 36) =59.897, p < 0.001; main effect of virus, F (1, 36) =14.293, p < 0.001). The
latency of P2 was comparable across the groups (Figure 16 I; all p > 0.05).
Further, compared to GFP-saline group, the P3 amplitude was larger in in tau-scopolamine and
GFP-scopolamine groups, but not in tau-saline group (Figure 15 L; three-way RM-ANOVA,
CStype × Virus × Drug interaction, F (1, 36) =0.328, p = 0.570; main effect of scopolamine, F (1, 36)
=15.813, p < 0.001). The latency of P3 was comparable across the groups (Figure 16 L; all p >
0.05).
Overall, the pattern suggests that while entorhinal HP-tau over-expression and scopolamine can
separately attenuate temporal auditory ERP amplitudes, simultaneous presentation of the two
conditions increases the level of ERP amplitudes higher than that of controls.
3.3.4 Visual stimulus-evoked ERPs in the frontal cortex
In GFP-saline group, the frontal visual ERPs displayed a pronounced P1 and N1 deflection, and a
smaller P2 deflection (Figure 17 A). The amplitude of P1 component in GFP-saline group was
comparable to tau-saline and GFP-scopolamine groups, but smaller than tau-scopolamine group
(Figure 18 A; three-way RM-ANOVA, CStype × Virus × Drug interaction, F (1, 36) =0.774,
p=0.385; Virus × Drug interaction, F (1, 36) =10.854, p=0.002; follow-up independent sample t-test,
GFP-saline vs. tau-saline, t= -0.173, p= 0.932; GFP-saline vs. tau-scopolamine, t=4.262, p < 0.001;
GFP-saline vs. GFP-scopolamine, t=-0.71, p =0.944). Moreover, the P1 amplitude in tau-
scopolamine was significantly larger than that in the tau-saline group (t (13.78) = 3.818, p = 0.002).
The latency of P1 was comparable across the groups (Figure 19 A; all p > 0.05).
Compared to GFP-saline group, the amplitude of N1 component was larger in tau-saline, GFP-
scopolamine, and tau-scopolamine groups respectively (Figure 18 D; three-way RM-ANOVA,
CStype × Virus × Drug interaction, F (1, 36) =1.395, p = 0.245; main effect of scopolamine, F (1, 36)
=28.250, p < 0.001). The latency of N1 was comparable across the groups (Figure 19 D; all p >
0.05).
59
Compared to GFP-saline, the amplitude of P2 component was larger in GFP-scopolamine and tau-
scopolamine groups, but similar in tau-saline group (Figure 18 G; three-way RM-ANOVA, CStype
× Virus × Drug interaction, F (1, 36) =0.694, p = 0.160; main effect of scopolamine, F (1, 36) =244.9,
p < 0.001; main effect of Virus, F (1, 36) =53.403, p < 0.001). The latency of P2 was comparable
across the groups (Figure 19 G; all p > 0.05).
Overall, the pattern suggests that while entorhinal HP-tau over-expression does not alter the levels
of frontal visual ERP amplitudes, scopolamine can increase frontal visual ERP amplitude levels
higher than that of controls. The effect of scopolamine is further potentiated in presence of
entorhinal HP-tau.
3.3.5 Visual stimulus-evoked ERPs in the parietal cortex
In GFP-saline group, the parietal auditory ERPs displayed a pronounced P1, N1, and P2 deflection
(Figure 17 B). The amplitude of parietal visual P1 component was comparable between the GFP-
saline group and tau-saline group, but was smaller in GFP-scopolamine and tau-scopolamine
groups compared to the GFP-saline group (Figure 18 B; three-way RM-ANOVA, CStype × Virus
× Drug interaction, F (1, 36) =0.110, p = 0.742; main effect of scopolamine, F (1, 36) =24.431, p <
0.001). Further, compared to GFP-saline group, the latency of P1 component was shorter in GFP-
scopolamine and tau-scopolamine groups (Figure 19 B; three-way RM-ANOVA, CStype × Virus
× Drug interaction, F (1, 36) =2.599, p = 0.116; main effect of scopolamine, F (1, 36) =9.638, p <
0.004).
Compared to GFP-saline group, the amplitude of N1 component was larger in GFP-scopolamine
and tau-saline group, but not in tau-scopolamine group. No effect of treatment was observed on
N1 amplitude or latency across all groups (Figure 18 E; all p > 0.00625; Figure 19 E; all p > 0.05).
Compared to GFP-saline group, the amplitude of P2 component was similar in tau-saline, but was
larger in GFP-scopolamine and tau-scopolamine groups (Figure 18 H; three-way RM-ANOVA,
CStype × Virus × Drug interaction, F (1, 36) =0.987, p = 0.327; main effect of scopolamine, F (1, 36)
=17.598, p < 0.001). The latency of P2 was comparable across the groups (Figure 19 H; all p >
0.05).
60
Overall, the pattern suggests that scopolamine results in an increase of the amplitude levels of
parietal visual ERPs, while entorhinal HP-tau over-expression seems not to have an effect.
3.3.6 Visual stimulus-evoked ERPs in the temporal cortex
The temporal visual ERP waveforms displayed a prominent P1, and a smaller N1 and P2
deflections (Figure 17 C). Statistical analyses did not reveal any significant differences between
the P1 amplitude of different groups (Figure 18 C; all p > 0.00625), nor any significant difference
were observed between the P1 latency of different groups (Figure 18 C; all p > 0.05).
Compared to GFP-saline group, the amplitude of visual N1 component was larger in tau-saline,
GFP-scopolamine, and tau-scopolamine groups (Figure 18 F; three-way RM-ANOVA, CStype ×
Virus × Drug interaction, F (1, 36) =0.630, p=0.433; main effect of scopolamine, F (1, 36) =9.898, p =
0.003). The latency of N1 was comparable across the groups (Figure 19 F; all p > 0.05).
Next, compared to GFP-saline group, the amplitude of visual P2 component was larger in tau-
saline, GFP-scopolamine, and tau-scopolamine groups (Figure 18 I; three-way RM-ANOVA,
CStype × Virus × Drug interaction, F (1, 36) =15.301, p<0.001; follow-up independent sample t-test,
GFP-saline vs. tau-saline, t= -9.005, p < 0.001; GFP-saline vs. tau-scopolamine, t=1.138, p=0.269;
GFP-saline vs. GFP-scopolamine, t=0.643, p=0.527) Moreover, the P2 amplitude in tau-saline was
significantly larger than that in the tau-scopolamine group (t (13.78) = 9.147, p < 0.001). The latency
of P2 component was comparable across the groups (Figure 19 I; all p > 0.05).
Overall, the pattern suggests that scopolamine as well as entorhinal HP-tau over-expression
increase the level of temporal visual N1 and P2 amplitudes respectively higher than that of
controls.
61
Figure 14. A depiction of grand average frontal, parietal, and temporal ERPs across treatment groups GFP-saline
(red), GFP-scopolamine (pink), tau-saline (dark blue) and tau-scopolamine (light-blue), during the paring of the
auditory CS with the US (ACS+; A-C) and during solo presentation of the auditory CS (ACS-; D-F). Statistical
analysis of the frontal ERP (A and D) revealed a significant effect of a three-way interaction between CStype
(ACs+/ACS-), Virus (entorhinal tau pathology), and Drug (cholinergic deficiency) on P1 and N1 components
(CStype × Virus × Drug; dark red), and a significant effect of a 2-way interaction between Virus and Drug on the
P2 component (Virus × Drug; purple). Examination of the parietal ERP (B and E) found a significant main effect
of Drug and a significant main effect of Virus on the P1 component and similarly on the N1 component (gray), a
significant effect of a three-way interaction on the P2 component (CStype × Virus × Drug; dark red), and a
significant effect of a 2-way interaction on the P3 component (Virus × Drug; purple). Over the temporal cortex,
the analyses found a significant effect of a 2-way interaction on the P1 and N1 components (Virus × Drug; purple),
a significant main effect of Drug and a significant main effect of Virus on the P2 component, and a significant
main effect of Drug on the P3 component (green).
Figure 14. Auditory ERPs during DTEBC paradigm.
GFP-Saline GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
ERP trace CStype × Virus × Drug
Virus × Drug
Main effect of Drug
Main effects of Drug and Virus
Treatment effect
No effect
Time from CS onset (s)
B C
E F D
A
Frontal Parietal Temporal
62
Figure 15. Line graphs display the effect of treatments on the mean amplitude of the auditory P1 (A-C), N1 (D-E),
P2 (G-I), and P3 (J-K) ERP components across frontal, parietal, and temporal cortices of GFP-saline (red), GFP-
scopolamine (pink), tau-saline (dark blue), and tau-scopolamine (light blue), during CS-US pairing (CS+) and solo
presentation of the CS (CS-). Statistical examination of the P1 mean amplitude (A-C) revealed a significant effect
of a three-way interaction between CStype (ACS+/ACS-), Virus (entorhinal tau pathology), and Drug (cholinergic
deficiency) on the frontal P1 (A; CStype × Virus × Drug; dark red), a significant main effect of Drug and a
significant main effect of Virus on the parietal P1 (B; gray), and a significant effect of a 2-way interaction on the
temporal P1 (C; Virus × Drug; purple). Similarly, analysis of N1 mean amplitude (D-F) revealed a significant effect
of a three-way interaction on the frontal N1 (D; CStype × Virus × Drug; dark red), and a significant main effect of
Virus on the parietal N1 (E; gray), and a significant effect of a 2-way interaction on the temporal N1 (F; Virus ×
Drug; purple). Examination of P2 mean amplitude (G-I) found a significant effect of a 2-way interaction on the
frontal P2 (G; Virus × Drug; purple), a significant effect of a three-way interaction on the parietal P2 (H; CStype ×
Virus × Drug; dark red), and a significant main effect of Drug and a significant main effect of Virus on the temporal
P2 (I; gray). Statistical analysis of P3 mean amplitude (J-L) found a significant effect of a 2-way interaction on the
parietal P3 (K; Virus × Drug; purple), and a significant main effect of Drug on the temporal P3 (L; green). Error
bars represent the standard error of the mean.
Figure 15. Interaction graphs of auditory ERP amplitude.
GFP-Saline
GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
Treatment groups
CStype × Virus × Drug
Virus × Drug
Treatment effects
Main effect of Drug
Main effects of Drug and Virus
63
Figure 16. Line graphs display the effect of treatments on the mean latency of the auditory P1 (A-C),
N1 (D-E), P2 (G-I), and P3 (J-K) ERP components across frontal, parietal, and temporal cortices of
GFP-saline (red), GFP-scopolamine (pink), tau-saline (dark blue), and tau-scopolamine (light blue),
during CS-US pairing (CS+) and solo presentation of the CS (CS-). Statistical analysis did not reveal
any significant effect of treatment in the latency of P1 component across different brain regions (A-C;
white). Examination of N1 mean latency revealed a significant main effect of Drug (cholinergic
deficiency) on the mean latency of parietal N1 (E; green). No treatment effects were observed on the
N1 mean latency over the frontal and temporal cortices (D and F; white), as well as the mean latency
of P2 and P3 components over the frontal, parietal, and temporal cortices (G-L; white). Error bars
represent the standard error of the mean.
Figure 16. Interaction graphs of auditory ERP latency.
GFP-Saline GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
Treatment groups
Main effect of Drug
Treatment effects
No effect
64
Figure 17. A depiction of grand average frontal, parietal, and temporal ERPs across treatment groups GFP-saline
(red), GFP-scopolamine (pink), tau-saline (dark blue) and tau-scopolamine (light-blue), during the paring of the
visual CS with the US (VCS+; A-C) and during solo presentation of the visual CS (VCS-; D-F). Statistical analysis
of the frontal ERP (A and D) revealed a significant effect of a 2-way interaction between Virus (entorhinal tau
pathology) and Drug (cholinergic deficiency) on the P1 component, a significant main effect of Drug on the N1
component, and a significant main effect of Drug and a significant main effect of Virus on the P2 component
(gray). Examination of the parietal ERP (B and E) found a significant main effect of Drug on the P1, N1, and P2
components (green). Over the temporal cortex, the analyses found a significant main effect of Drug on the N1
component (green), and a significant effect of a three-way interaction between CStype (VCS+/VCS-), Virus, and
Drug on the P2 component (CStype × Virus × Drug; dark red).
Figure 17. Visual ERPs during DTEBC paradigm.
Time from CS onset (s) CStype × Virus × Drug
Virus × Drug
Treatment effect
No effect
Main effect of Drug
Main effects of Drug and Virus
GFP-Saline GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
ERP trace
Frontal Parietal Temporal
65
Figure 18. Line graphs display the effect of treatments on the mean amplitude of the visual P1 (A-C), N1 (D-E), P2
(G-I), and P3 (J-K) ERP components across frontal, parietal, and temporal cortices of GFP-saline (red), GFP-
scopolamine (pink), tau-saline (dark blue), and tau-scopolamine (light blue), during CS-US pairing (CS+) and solo
presentation of the CS (CS-). Statistical examination of the P1 mean amplitude (A-C) revealed a significant effect of a
2-way interaction between Virus (entorhinal tau pathology) and Drug (cholinergic deficiency) on the frontal P1 (A;
Virus × Drug; purple), a significant main effect of Drug on the parietal P1 (B; green), and no significant effect on the
temporal P1 component (C; white). Analysis of N1 mean amplitude (D-F) revealed a significant main effect of Drug
on the frontal N1 (D; green), a significant effect a 2-way interaction between CStype (CS+/CS-) and Drug on the
parietal N1 (E; CStype × Drug; pink, and a significant main effect of Drug on the temporal N1 (F; green). Examination
of P2 mean amplitude (G-I) found a significant main effect of Drug and a significant main effect of Virus on the frontal
P2 (G; gray), a significant main effect of Drug on the parietal P2 (H; green), and a significant effect of a three-way
interaction on the temporal P2 (I; CStype × Virus × Drug; dark red), No treatment effects were observed on the P3
mean amplitude over the frontal and temporal cortices (J-L; white). Error bars represent the standard error of the mean.
Figure 18. Interaction graphs of visual ERP amplitude.
GFP-Saline GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
Treatment groups
Main effects of Drug and Virus
CStype × Virus × Drug
Virus × Drug
Treatment effect
No effect
Main effect of Drug
66
Figure 19. Line graphs display the effect of treatments on the mean latency of the visual P1 (A-C), N1 (D-E), P2
(G-I), and P3 (J-K) ERP components across frontal, parietal, and temporal cortices of GFP-saline (red), GFP-
scopolamine (pink), tau-saline (dark blue), and tau-scopolamine (light blue), during CS-US pairing (CS+) and
solo presentation of the CS (CS-). Statistical analyses P1 mean latency revealed a significant main effect of Drug
(cholinergic deficiency) on the mean latency of parietal P1 component (E; green). No treatment effects were
observed on the P1 mean latency over the frontal and temporal cortices (A and C; white), as well as the mean
latency of N1, P2, and P3 components over the frontal, parietal, and temporal cortices (D-L; white). Error bars
represent the standard error of the mean.
Figure 19. Interaction graphs of visual ERP latency.
GFP-Saline GFP-Scopolamine
Tau-Saline
Tau-Scopolamine
Treatment groups
Main effect of Drug
Treatment effects
No effect
67
3.4 SVM-based classification of ERPs
Figure 20 (A-F) show the classification accuracy of SVM for GFP-saline, tau-saline, GFP-
scopolamine, and tau-scopolamine based on different combination of extracted ERP features.
When all of the 96 ERP features were incorporated in the classification procedure, the
classification accuracy of the SVM was 100% for GFP-saline, 100% for tau-saline, 100% for GFP-
scopolamine, and 90% for tau-scopolamine (Figure 20 A; all features). The accuracy for all groups
was significantly better than what was expected by chance (binominal test, p < 0.05). To identify
which combination of features yield the best classification accuracy, we then repeated the SVM
analysis by using a subset of 96 features.
Examination of the effect of stimulus modality on SVM’s classification accuracy revealed no
significant difference between the auditory versus visual stimulus. With either 48 auditory or 48
visual ERP features SVM was able to detect all the treatment groups with high accuracy (Figure
20), and the accuracy for all groups was significantly better than what was expected by chance
(binominal test, p < 0.05).
Next, the effect of CS-US contingency on SVM classification accuracy was investigated.
Accuracy of SVM did not differ based on CS-US contingency (Figure 20 C). Specifically, the
accuracy obtained with 48 ERP features during CS+ was comparable with the accuracy obtained
during the CS-, and the accuracy for all groups was significantly better than what was expected by
chance (binominal test, p < 0.05).
Furthermore, the discriminatory powers of the two ERP peak properties, amplitude and latency,
were compared. By using 48 ERP amplitude features SVM was able to detect all the groups, but
with 48 ERP latency features only GFP-scopolamine group was detected. The accuracy was
significantly better than what was expected by chance (Figure 20 D; binominal test, p < 0.05).
Next, the discriminatory powers of ERP data from frontal, parietal, and temporal cortices were
compared. The accuracy of the classifier was comparable across different brain regions, and was
significantly better than what was expected by chance (Figure 20 E; binominal test, p < 0.05). This
suggests that features from one brain regions are sufficient for detecting different groups, among
them features from the temporal cortex allowed the classifier to detect the groups with 100%
accuracy.
68
Moreover, extracted features from P1, N1, P2, and P3 ERP components were tested separately to
find the component that yielded the highest classification accuracy. Majority of the ERP
components allowed the classifier to detect the pathologies accurately and significantly better than
what was expected by chance (binominal test, p < 0.05), with exception of P2 component for tau-
scopolamine group, and P3 component for GFP-scopolamine group (Figure 20 F).
Together, the results indicate that different pathologies such as accumulation of HP-tau protein in
the entorhinal cortex, or scopolamine-induced deficiency of the central cholinergic system, induce
abnormalities in the pattern of brain’s ERPs that are distinct from each other. These abnormal
patterns are observed mainly as alterations in the level of the peak amplitude of the ERP
components and can be utilized by machine learning classifiers such as a support vector machine
to detect the type of pathology present (Figure 21).
69
Figure 20. SVM classifier accuracy.
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
Light Tone
A) All features
All Features B) Stimulus modality
CS+
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
CS-
C) CS-US contingency
0 20 40 60 80
100 120
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
D) ERP peak properties
Latency Amplitude
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
2 × 2 × 2 × 3 × 4 = 96 features
1 × 2 × 2 × 3 × 4 = 48 features
2 × 1 × 2 × 3 × 4 = 48 features
2 × 2 × 1 × 3 × 4 = 48 features
Tau-Saline Tau-Scopolamine
GFP-Saline GFP-Scopolamine
70
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
E) Brain regions
Frontal
Temporal
Parietal
F) ERP components
N1 P1
P2 P3
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
2 × 2 × 2 × 1 × 4 = 32 features
Cla
ss
ific
ati
on
ac
cu
racy
(%
)
2 × 2 × 2 × 3 × 1 = 24 features
Figure 20. Bar graphs display SVM’s
classification accuracy (CA) for GFP-saline
(Gsal; red), tau-saline (Tsal; blue), GFP-
scopolamine (Gsco; pink), and tau-
scopolamine (Tsco; light blue) with different
combinations of extracted ERP features as
indicated in the charts (Blue, selected features;
Yellow, compared features). A) The CA with
all 96 features was 100% for Gsal, 100% for
Tsal, 100% for Gsco, and 90% for Tsco.
B) Across stimulus modalities, auditory stimulus (tone) yielded a CA of 100% for Gsal, 100% for Tsal, 100% for Gsco, and
90% for Tsco, and visual stimulus (light) yielded a CA of 100% for Gsal, 100% for Tsal, 100% for Gsco, and 90% for Tsco.
C) Features extracted during CS-US pairing (CS+) resulted a CA of 100% for Gsal, 100% for Tsal, 100% for Gsco, and 90%
for Tsco, and features extracted during solo CS presentation (CS-) yielded a CA of 90.91% for Gsal, 100% for Tsal, 100%
for Gsco, and 90% for Tsco. D) When comparing the properties of the ERP peak, ERP peak amplitude achieved a CA of
100% for Gsal, 100% for Tsal, 100% for Gsco, and 90% for Tsco, while ERP peak latency yielded an accuracy level of
45.4% for Gsal, 0 for Tsal, 81.8% for Gsco, and 30% for Tsco. E) Among different brain regions, features from frontal ERP
yielded a CA of 100% for Gsal, 100% for Tsal, 90.1% for Gsco, and 80% for Tsco, while parietal ERP features achieved a
CA of 81.8% for Gsal, 75% for Tsal, 81.9% for Gsco, and 80% for Tsco, and temporal ERP features resulted in a CA of
100% for Gsal, 100% for Tsal, 100% for Gsco, and 100% for Tsco. F) When comparing separate ERP components, CA with
P1 component was 81.2% for Gsal, 100% for Tsal, 81.2% for Gsco, and 90% for Tsco. Features from N1 component also
yielded a CA of 100% for Tsal and 90% for Tsco, while increasing the CA for Gsal and Gsco to 90.9% and 100% respectively.
The CA achieved with features from P2 component was 100% for Gsal, 100% for Tsal, 100% for Gsco, and 50% for Tsco,
and with features from P3 component was Gsal, 75% for Tsal, 45.4% for Gsco, and 60% for Tsco.
Tau-Saline Tau-Scopolamine
GFP-Saline GFP-Scopolamine
71
Figure 21. Summary of SVM classifier accuracy.
Figure 21. The heat map summarized the classification accuracy (CA) results (dark blue 0-100%
dark red) of the SVM classifier for each treatment groups (GFP-saline, Tau-saline, GFP-
scopolamine, and Tau-scopolamine; column) with different combinations of test features (rows).
For stimulus modality, tone-CS and light-CS had comparable CAs. For CS contingency, reinforced
CS with the US (CS+) and the non-reinforced CS (CS-) had similar CAs. Between the properties
of the ERP peak, amplitude displayed an exceedingly better accuracy than latency.
Classification results with features from frontal, parietal, and temporal cortices were comparable.
Within different ERP components, P2 and P3 failed to detect Tau-scopolamine and GFP-
scopolamine respectively.
72
Chapter 4 Discussion
4.1 Comparison of histological changes with AD staging
Based on our histological analysis we observed that HP-tau was accumulated in the entorhinal
cortex, but was absent in two of entorhinal cortex efferent regions, hippocampus and prelimbic
area. Together these results suggest that in our model we successfully induced tau pathology that
corresponds to the transentorhinal stage of preclinical AD (427).
Although systemic injection of scopolamine at the dose used in this study allowed us to induce
cholinergic deficiency without memory impairment, the mechanism that induced cholinergic
deficiency in our model was not the same as the mechanism observed in preclinical stage of AD.
In preclinical AD, a presynaptic cholinergic deficit is caused by the reduction in choline
acetyltransferase activity which results in a lowered level of acetylcholine production (428),
whereas in our model, scopolamine induces a postsynaptic cholinergic deficit by blocking
muscarinic receptors. Another difference is that the reduction in cholinergic input during
preclinical AD occurs progressively over the years, but scopolamine induces an acute deficiency
in cholinergic input.
4.2 Effects of preclinical AD pathology on associative learning and reversal learning
Using an associative memory paradigm known as differential trace eyeblink conditioning
(DTEBC), we show that associative learning and reversal learning is not impaired in rats with
entorhinal HP-tau over-expression, cholinergic deficiency, or both pathologies in conjunction. We
tested associative learning during a ten-day acquisition phase, in which rats with brain pathology
and the control rats acquired a preemptive eyeblink conditioned response (CR) to the reinforced
conditioned stimulus (CS+) but not to the neutral CS (CS-). Next, we assessed reversal learning
by switching the contingency between CS and the unconditioned stimulus (US), and observed that
73
rats with pathology, similar to control rats, were able suppress their CR to the previously-
reinforced CS and acquire CR to the new CS+.
The absence of cognitive deficits observed in this study is consistent with other studies that
demonstrated that preclinical AD pathologies do not produce cognitive deficits. For example, a
study in mice with accumulated HP-tau in EC and hippocampus reported that pathology at this
stage is not robust enough to cause deficits in spatial navigation memory (429). Another study
with clinically normal adults who exhibited biomarker evidence of Aβ deposition did not find any
impairment in associative memory examined with cued recall test. Our observation on the
behavioral effects of the preclinical AD pathology in rats is in parallel with previous studies in
rodents and humans, which report that AD neuropathology in the preclinical stage of the disease
is asymptomatic (429-431).
4.3 Mechanisms underlying ERP abnormality and susceptibility of different cortical regions to features of preclinical AD pathology
During differential trace eyeblink conditioning, we also recorded event related potentials during
the trace interval between the CS and US. Our ANOVA-based comparison of ERP peak amplitude
and latency showed that entorhinal HP-tau over-expression significantly decreased the amplitude
of frontal auditory P2, parietal auditory P1, N1, and P3, and temporal auditory P1, N1, and P2
components, whereas it significantly increased the amplitude of frontal and temporal visual P2
component.
In addition, cholinergic deficiency significantly decreased the amplitude of parietal auditory P1,
N1, and P3, parietal visual P1, and the temporal auditory P1 and N1 components. However,
cholinergic deficiency significantly increased the amplitude of the temporal auditory P2 and P3,
temporal visual N1, and frontal visual N1 and P2 components. Further, the latency of parietal
auditory N1 and visual P1 were significantly as a result of cholinergic deficiency.
Interestingly, we observed abnormal parietal auditory P3 component rats with entorhinal HP-tau,
cholinergic deficiency, and conjunction of both pathologies (Figure. 14 B). Similar findings have
been reported in human oddball paradigm studies with probable AD and MCI patients where the
74
amplitude of auditory P3 component was reduced compared to age-matched controls (432, 434,
435). Our finding suggests that pathological features of AD appear to affect ERP components
similarly in rats and humans.
Further, based on the number of affected ERP components by each pathological condition in
separate brain regions, we observed that temporal ERPs were mainly affected by the entorhinal
HP-tau over-expression, parietal ERPs were mostly impacted by cholinergic deficiency, and
frontal ERPs were largely affected by the conjunction of both conditions. Examination of ERP
features from each brain region with the classifier however, revealed that ERPs from one cortical
area are capable of accurately detecting different types of pathologies (Figure. 20 E). These
findings suggest that, different brain regions may be affected by AD pathologies at different levels,
and that the within one cortical region different pathologies affect ERPs with substantially distinct
pattern.
4.3.1 HP-tau-induced synaptic dysfunction and cortical hyperexcitability
The temporal ERPs are mostly affected by HP-tau over-expression likely because HP- tau results
in loss of dendritic spines, disrupting axonal transport, and synaptic anchoring which ultimately
leads to synaptic dysfunction of entorhinal neurons (436-438). Indeed, many studies have
implicated HP-tau in synaptic dysfunction and cortical hyperexcitability (439, 440). It has been
shown that localization of HP-tau to the spines of hippocampal pyramidal cells in 4-month-old
rTg4510 mice can initiate spine loss and dendritic regression by decreasing α-amino-3-hydroxy-
5-methyl-4-isoxazolepropionic acid receptor (AMPA)-mediated synaptic currents through
removal of glutamate receptor 1 (GluR1) AMPA receptors from spines (441, 442). Loss of spine
density reduces the dendritic diameter of the presynaptic cortical neuron and attenuates the
frequency and the amplitude of the synaptic current moving through the dendrites (443).
Further, HP-tau can reduce synaptic synchrony by impairing axonal transport and synaptic
anchoring (444-449). Reduction of synaptic synchrony together with the attenuation of frequency
and amplitude of the synaptic current cause insufficiency of synaptic input to the postsynaptic
neurons as observed in the hippocampus of transgenic mouse with P301S mutation (450). This
75
may explain the attenuated peak amplitude of the ERP components in rats with entorhinal HP-tau
over-expression (Figure. 14).
As the pathology progress, increasing number of presynaptic cortical neurons are lost, leaving a
significant population of postsynaptic neurons deafferented (451). It is speculated that the loss of
afferent input is compensated by the surviving neurons in the 9-month-old rTg4510 mice through
increased excitability, as observed by elevated frequency of glutamatergic spontaneous excitatory
postsynaptic currents (sEPSCs) and increased amplitude of spontaneous inhibitory postsynaptic
currents (sIPSCs), and sprouting of new axonal buttons and formation of new synapses (452).
These compensatory homeostatic mechanisms may also account for the long preclinical stage of
the AD, during which maintaining network stability supports cognitive functions until the
pathology affects the compensatory mechanisms and results in a severe network dysfunction (453).
4.3.2 Cholinergic deficiency and the hippocampus
In addition to the abnormalities in temporal ERPs, we detected that parietal ERPs are particularly
sensitive to cholinergic deficiency. The sensitivity of parietal ERPs to cholinergic deficiency in
rats may be attributed to disruption of neuronal activity in the underlying hippocampal structures,
which receives dense cholinergic projections from the medial septal nucleus (454-456).
The pivotal role of cholinergic system in cognitive processes such as working memory and
attention has been well established through lesion and pharmacological studies in rodents and
nonhuman primates (457-461). It is known that acetylcholine modulates synaptic plasticity in
hippocampus and facilitates the firing of hippocampal pyramidal neurons through muscarinic
receptors (462-465). Additionally, studies in rats have shown that blocking of muscarinic receptors
by scopolamine inhibits long-term potentiation (LTP) in the dentate gyrus and the CA1 regions
(466, 467) and that muscarinic agonists enhanced hippocampal LTP (468).
Moreover, cholinergic modulation may influence cognitive functions through the processing of
sensory information, in which cholinergic neurons project to many cortical targets to facilitate
neural sensory processing by enhancing signal-to-noise ratio at the synaptic level (469-474).
Therefore, the effect of cholinergic deficiency on parietal ERP may also be an outcome of
76
disrupted cholinergic-dependent modulation of mechanisms that take part in processing of inputs
to the hippocampus (475-478).
4.3.3 HP-tau, cholinergic deficiency, and the abnormal frontal ERPs
Unlike temporal and parietal ERPs, frontal ERPs are sensitive to the conjunction of entorhinal HP-
tau and cholinergic deficiency. This sensitivity may be due to disruption of input from the
hippocampus, entorhinal cortex, and the basal forebrain cholinergic projections to the frontal
cortex.
It is known that medial prefrontal cortex (mPFC) is involved in a number of cognitive functions
such as learning and memory (479-481) and is believed to mediate its function through multiple
connections with limbic and sensory cortical structures in the parietal and temporal cortices (482-
485).
The membrane potential of the mPFC pyramidal neurons in rats oscillates from an inactive
hyperpolarized state to a depolarized state at which action potentials are fired (486, 487).
Hippocampal contribution to regulation of mPFC activity is believed to arise from modulation of
this bistable state by a combination of excitatory and inhibitory inputs from the hippocampus to
the mPFC (486, 489). Further, studies with retrograde tracers have shown that entorhinal cortex
sends nonreciprocal projections from its layer III and V that synapse onto infralimbic and
prelimbic mPFC respectively (490). In vivo recordings of prelimbic and infralimbic neurons in
rats has shown that stimulation of entorhinal cortex results in a major inhibition of mPFC neurons
and that this response is not altered by disruption of hippocampal projections to mPFC (491).
Contribution of cholinergic deficiency to aberrant frontal ERPs may be due interruption of
cholinergic-dependent modulation of local interneurons in the frontal cortex. It is known that
medial and lateral PFC are heavily innervated by projections from the BF cholinergic and
GABAergic neurons (492, 493).
The majority of cholinergic neurons project to somatostatin (SOM)-expression interneurons, while
the remaining contact vasoactive intestinal polypeptide (VIP)-expressing interneurons (494). The
SOM-expressing interneurons are depolarized by activation of muscarinic receptors and inhibit the
77
dendrites of the pyramidal cells, whereas VIP interneurons provide disinhibitory control by
preferentially targeting and inhibiting other interneurons upon depolarization (494, 495).
It appears that the cholinergic projections via the SOM-expressing and VIP interneurons are able
to be able to differentially modulate the activity of frontal pyramidal cells. However, as the
majority of cholinergic projections are to the SOM-expressing interneurons, blockade of
muscarinic receptors may disinhibit pyramidal cells in the frontal cortex and increase the level of
observed frontal ERP.
Further, it has been shown that the BF GABAergic neurons contact the pyramidal cells in the
frontal cortex by direct projections and also indirectly via the parvalbumin (PV) interneurons (494-
496). Recently, it has been shown that a subset of BF neurons fire a phasic burst response that
precedes the frontal ERP activity by a few milliseconds every time a behaviorally-relevant
stimulus is presented, and that the change in the amplitude of the bursting response to relevant
versus irrelevant stimuli is highly comparable to the change in the frontal ERP amplitude (498).
This bursting firing pattern has been attributed to the activity of the GABAergic neurons, and it
has been suggested that they initiate the activity of the pyramidal cells by disinhibition via the PV
interneurons (499-501).
4.4 Use of ERP features as a detection tool for early brain pathology
Next, we asked whether different types of pathologies could be detected from ERPs. We tested all
of the extracted ERP features with the support vector machine (SVM) classifier and observed that
all three types of pathologies and the control group were detected with high accuracy (Figure. 20
A). SVM-based classification of ERPs have been successfully used in the past to examine the
neurophysiological state of the brain in humans (502, 503). Consistent with our finding,
application of SVM classifier to extracted ERP features in humans have shown promising results
for early diagnosis of AD (504). Similarly, other studies have used the SVM-based classification
on magnetic resonance imaging (MRI) results to differentiate between normal and AD patients,
and to predict conversion from MCI to AD (505, 506). In addition, SVM classifier has been applied
to combined EEG-MRI data for differential diagnosis of AD and dementia with Lewy bodies
78
(507). Our finding in addition to these evidence indicate that ERPs indeed contain diagnostically
useful information.
4.4.1 Effect of stimulus modality and contingency on ERP-based detection of pathology
By using different stimulus modalities, we examined which of auditory or visual stimuli are better
for detecting pathology. A comparison between the total number of affected auditory ERPs versus
the visual ERPs revealed that a higher number of auditory ERP components than visual ERP
components were affected by entorhinal HP-tau over-expression alone, and in conjunction with
cholinergic deficiency. This finding suggests that auditory ERPs may be more susceptible to
neuropathological changes in the brain than the visual ERPs. The likely reason is that the sources
of the auditory ERPs in the temporal lobe are closer to the site of pathology in the entorhinal cortex
than the sources of the visual ERPs, which are thought to be located at the inferior occipital area
(508-511).
Nonetheless, when the extracted auditory and visual ERP features were tested separately with the
SVM classifier the outcome was comparable between different modalities and all pathologies were
detected with high accuracy (Figure. 20 B), indicating that the discriminatory power of auditory
ERPs is comparable to visual ERPs. Consistent with our observation, it has been reported that
auditory ERPs have a comparable sensitivity scores to visual ERPs for diagnosis of individuals
with early-stage AD, and for differentiation of progressive MCI from the stable form (512-515).
We also observed that the overall pattern of ERPs in each brain region was comparable between
CS+ and CS- conditions for auditory and visual stimulus. (Figure. 14 & 17). Subsequently, we
separately tested ERP features from CS+ and CS- conditions with the classifier, and observed that
the classification accuracy was comparable between the data sets and that all pathologies were
detected with high accuracy (Figure. 20 C). This observation suggests that CS-US contingency
does not substantially alter the overall pattern and consequently the discriminatory power of the
cortical ERPs.
Although distinct properties of ERPs including, amplitude and latency, are dependent on the
specific CS-US contingency implemented, some ERP components can overcome these
79
contingency-dependent changes (516). Specifically, it appears that CS-US contingency has a
stronger effect on ERP components with longer latencies (P2, P3) than the earlier ERP components
(P1, N1) in rats, most likely because longer time windows allow for greater level of impact form
the neuromodulators, such as the cholinergic projections (517, 518).
Additionally, it is possible that unlike allocortical structures such as the entorhinal cortex and the
hippocampus that are involved in associative learning and formation of discrete conditioned (519-
521) the neocortical structures, where cortical ERPs are thought to be originating (522) are more
involved with perceptual aspect of learning than with stimulus-response adaptations (523).
4.4.2 Susceptibility of ERP properties to preclinical AD pathology
We observed that entorhinal HP-tau, cholinergic deficiency, and both mainly affected the peak
amplitude and not the peak latency of the ERP components. Using ERP amplitude features, SVM
classifier was able to detect all pathologies with high accuracy; however, when the ERP latency
features was tested, only rats with cholinergic deficiency were detected (Figure. 20 D). Our
findings suggest that among ERP properties, amplitude has stronger discriminatory power and is
more accurate than latency for detection of AD pathology.
Currently, due to conflicting reports from different studies there is no consensus over which ERP
property is the most sensitive for detection of AD pathology. Recent ERP studies in AD have
mainly focused on the P3 component, which is thought to reflect brain’s information processing
when attentional and memory systems are activated (524). Some studies have reported a significant
reduction of P3 peak amplitude (525, 526) while others observed a significant prolongation of P3
peak latency in AD patients (527-529).
It is believed that amplitude and latency of P3 component respectively reflect the level of
attentional resources present (530), and the speed of stimulus detection and classification (531).
Therefore, reduction of P3 amplitude can occur when more resources are allocated for processing
of tasks with increasing attentional demands, or when fewer resources are available due to neuronal
death. Similarly, prolongation of P3 latency can be a consequence of disrupted communication
between neural networks that has occurred because of axonal damage and synaptic dysfunction
(532, 439, 440). Since synaptic dysfunction, damage to axonal transport, and neuronal death are
80
all part of AD neuropathology, variability in their combined effect on brain electrophysiology
across AD patients may explain the conflicting reports on the observed changes on P3 components.
Many studies in humans have also shown that increased attention and arousal level results higher
P1, N1, and P2 peak amplitudes (533-539), while other studies have either observed a reduction
in the latency or no effect at all ( 540-544). While our findings supports the idea that pathology-
induced changes in the level of attentional resources may reflect as changes in the ERP amplitude,
we did not find any evidence to suggest P3 is more reliable than the earlier components for
detecting these alterations. Other factors that could give rise to current conflicting observations are
discrepancy in recording procedures and differences in experimental paradigms across studies
(545). Additionally, factors such as multiple intra-component peaks and inter-personal variation
in topographic distribution can further complicate the measurement and analysis of ERP peak
amplitude and peak latency (546).
4.4.3 Detection of AD pathology with a single ERP component
Although different pathological features of AD affected several of the same ERP components in
one cortical region, there were specific ERP components that were altered only in one of the
pathological conditions. We specifically observed that temporal visual P2 component was only
affected by entorhinal HP-tau over-expression, while parietal visual P1, frontal and temporal visual
N1, and temporal auditory P3 were affected by cholinergic deficiency, and frontal visual and
auditory P1 was affected by conjunction of both conditions. In parallel, we observed that combined
auditory and visual features from one ERP component were sufficient for detection of all
pathologies with the classifier, except for P2 features that did not detect HP-tau over-expression
in conjunction with cholinergic deficiency, and P3 features that did not identify rats with
cholinergic deficiency (Figure. 20 F).
Together these findings suggest that, although entorhinal HP-tau over-expression and cholinergic
deficiency affect the same ERP component in one cortical area, the pattern by which these
pathologies affect the component (i.e., increase or decrease of amplitude and the magnitude of
change) is distinct. This is consistent with our hypothesis that different pathological features of
preclinical stage AD induce distinct abnormal ERP patterns, and that this distinct pattern allows
81
the classifier to detect and differentiate between pathologies using features from one ERP
component.
4.5 Summary and conclusion
Overall, we found that features of preclinical AD pathology cause ERP abnormalities in absence
of detectable memory impairment. Consistent with preclinical AD patients, rats with preclinical
AD pathology did not show any deficits in associative memories (Figure. 12 & 13) despite showing
significant alterations in ERP activity (Figure. 14 & 15).
By testing ERP data with a machine learning algorithm, we show that separate pathologies can be
detected from ERP features independent of stimulus modality (Figure. 20 A & B), indicating that
different pathological features manifest themselves as distinct ERP abnormalities in the brain.
These abnormalities are mainly exhibited as changes in the ERP peak amplitude (Figure. 20 D)
and are not influenced by CS-US contingency (Figure. 20 C). ERPs from one brain region are
sufficient for detection of all pathological feature, among which temporal ERPs appear to be more
accurate that frontal and parietal ERPs (Figure. 20 E). Moreover, all ERP components can detect
separate pathological features, except entorhinal HP-tau over-expression in conjunction with
cholinergic deficiency for P2 component, and cholinergic deficiency for P3 component.
Together our findings indicate that, ERPs may be a useful non-invasive and inexpensive tool for
detection of Alzheimer’s disease pathology in asymptomatic elderlies, and that SVM-based
classification may enhance the detection of aberrant ERPs during a differential associative learning
paradigm.
4.6 Limitations
It must be noted that the expression of mutated tau was not specific to neurons, because the chicken
β-actin promoter in rAAV9 recombinant viral vector is ubiquitously found in neurons as well as
glial cells (411). Thus, the observed effects of HP-tau over-expression may not solely be a result
82
of accumulation of mutated tau in neural population of the entorhinal cortex. Moreover, delivery
of the viral vector via the infusion cannula damages the brain tissue as the cannula is lowered into
the target coordinate, which may cause inflammation around the injection tract and cause an
interference with the activity of neurons unrelated to the effects of HP-tau.
Further, due to limited number of recording channels we were restricted in number of brain regions
we could record. Since EEG is reflective of brain’s global activity originating from multiple
cortical and subcortical sources, recording from the occipital cortex in addition to frontal, parietal,
and temporal cortices could have provided a more complete picture of how the global brain
network is affected by different pathological features of AD.
Next, to induce cholinergic deficiency in the brain, we systemically administrated muscarinic
receptor blocker, scopolamine hydrobromide (0.05 mg/kg), which inhibited central as well as the
peripheral muscarinic receptors. However, inhibition of peripheral muscarinic receptors by
scopolamine at doses 0.03 mg/kg and 1.0 mg/kg, have been reported to have no effects on
associative learning in rats and rabbits (398). Further, pharmacological blockade of muscarinic
receptors in our manipulation induced a reversible acute cholinergic deficiency, which is not
consistent with gradual attenuation of cholinergic input thought the course of preclinical AD.
Further, each of our experimental models only capture one of many pathologies in preclinical
stage. Therefore, it is easier to detect pathology than the reality in which multiple types of
pathology may induce a complexed mixture of ERP abnormality.
Lastly, although the conditioning chambers used in this study were sound attenuated, there was
always the possibility of environmental noise distracting the animals during the conditioning
sessions.
4.7 Future Direction
In this study we showed that pathological features of preclinical stage AD, including entorhinal
HP-tau over-expression, cholinergic deficiency, or both, induce distinct abnormal ERP patterns
during DTEBC to a degree which is sufficient for an accurate prediction of the type of brain
pathology based on the ERP patterns.
83
Next series of experiments need to investigate whether the other pathological feature of preclinical
AD, cortical amyloidosis, induces memory dysfunction and aberrant ERP activity. As such, will
record EEG activity from rats with cortical amyloidosis, cortical amyloidosis with cholinergic
deficiency, and cortical amyloidosis with entorhinal HP-tau overexpression while the animals
receive training in tone-light DTEBC paradigm.
In groups of rats examined in this study and the other groups that are proposed for the next series
of experiments, we will track ERPs transactionally to observe how ERP abnormality changes as
pathology is progressed. Moreover, we will examine the amplitude of specific frequency band of
oscillations and synchronization of oscillations between the recording sites in these groups of rats.
Additionally, we will implement the same conditioning paradigm in transgenic rats with multiple
pathological features to examine how the complex interaction of multiple pathologies affects
ERPs. Specifically, potentially useful models is the TgF344-AD rat model, which expresses
mutant human APP and PS1 gene, and exhibits age-dependent cortical amyloidosis followed by
tau pathology, gliosis, and neuronal death in the hippocampus and cortex (547). In addition, a rat
model with a strong risk factor for sporadic AD, human apolipoprotein e4 would be useful to
expand our findings to the majority of patients who do not carry causative mutations for AD (548).
At Last, since the DTEBC paradigm in animals can be applied to human subjects without any
major modification, these results hold a significant translational potential. To test our approach in
patient population we propose several steps that may help with successful implementation of this
strategy.
First, we suggest that all testing centers evaluate patients’ cognitive function using the same test,
such as the Mini-Mental State Examination (MMSE), and that they utilize one standardized EEG
recording protocol and acquisition setup up to minimize discrepancy in clinical assessment and the
obtained ERP data. Next, we suggest tone-light DTEBC with an eyeblink-eliciting air puff to be
used as the conditioning paradigm during EEG recording, and that the ERP data form patients be
compared with ERPs from healthy individuals who have been matched for age, sex, and education.
To maximize our chances of detecting AD at the preclinical stage, we suggest forming an ERP
data repository by compiling the obtained ERP data from all testing centers. The data in this
repository will be organized as three main categories of ERPs from healthy controls, individuals
84
at risk of developing AD, and the diagnosed patients; and within each category the data is indexed
based on the MMSE score. Once the number of ERP data samples in the repository reach a
threshold that can train a machine learning algorithm which has a high classification accuracy (e.g.
95% <), we can test the ERP data of asymptomatic individuals using the trained machine classifier.
85
References
1. Hippius, H. (2003). The discovery of Alzheimer’s disease, Dialogues in Clinical
Neuroscience. 5, 101-108.
2. Lage, J.M.M. (2006). 100 Years of Alzheimer's disease (1906-2006)., J Alzheimers Dis, 9, 15.
3. Braak, H., Braak, E. (1991). Neuropathological stageing of Alzheimer-related changes. Acta
Neuropathol, 82, 239-259.
4. Braak, H., Braak, E. (1997). Frequency of stages of Alzheimer-related lesions in different age
categories. Neurobiol Aging, 18, 351–357.
5. Mirra, S.S., Heyman, A., McKeel, D., Sumi, S.M., Crain, B.J., Brownlee, L.M., Vogel, F.S.,
Hughes, J.P., van Belle, G. & Berg, L. (1991). The Consortium to Establish a Registry for
Alzheimer's Disease (CERAD). Part II. Standardization of the neuropathologic assessment.
Neurology, 39, 1159-65.
6. Khachaturian, Z.S. (1985). Diagnosis of Alzheimer's disease, Arch Neurol, 42, 11..
7. Beach, T.G. (1997). Cholinergic fiber loss associated with diffuse plaques in the non-
demented elderly, the preclinical stage of Alzheimer's disease? Acta Neuropahtology 93(2),
146-53.
8. The Ronald and Nancy Reagan Research Institute of the Alzheimer’s Association and the
National Institute on Aging. (1998). Consensus report of the Working Group on, Molecular
and Biochemical Markers of Alzheimer’s Disease. Neurobiol Aging., 19 (2), 109–116.
9. Alzheimer Society Canada. (2016). Dementia numbers in Canada. Received from
http,//www.alzheimer.ca/en/About-dementia/What-is-dementia/Dementia-numbers
10. Price, J.L., Morris, J.C. (1999). Tangles and plaques in nondemented aging and preclinical
Alzheimer’s disease. Ann Neurol, 45(3), 358–368.
11. Bird, T.D. (2015). GeneReviews, Seattle, WA, University of Washington, Seattle.
86
12. Alzheimer, A. (1907). Über eine eigenartige Erkrankung der Hirnrinde. Allgem Z Psychiatr-
Gerich, 64, 146-148.
13. Lage, J.M.M. (2006). 100 Years of Alzheimer's disease (1906-2006), J Alzheimers Dis, 9, 15.
14. Crouch, P.J. (2007). The modulation of metal bio-availability as a therapeutic strategy for the
treatment of Alzheimer's disease, FEBS J, 274, 3775.
15. Kamer, A., Craig, R., Dasanayake, A., Brys, M., Glodzik-Sobanska, L. & de Leon, M.
(2008). Inflammation and Alzheimer's disease, Possible role of periodontal diseases,
Alzheimers Dem, 4, 242-250.
16. Itzhaki, R.F. & Wozniak, M.A. (2004). Alzheimer's disease, the neuroimmune axis, and viral
infection. Journal of Neuroimmunology, 156, 1-2.
17. Turner, P.R. (2003). Roles of amyloid precursor protein and its fragments in regulating
neural activity, plasticity and memory, Prog Neurobiol, 70, 1.
18. Vasto, S., Candore, G. (2008). Inflammation, genes and zinc in Alzheimer's disease, Brain
Res Rev, 58, 96.
19. Priller, C., Bauer, T. (2006). Synapse formation and function is modulated by the amyloid
precursor protein. J Neurosci, 26, 7212.
20. Priller, C., Mitteregger, G. (2009). Excitatory synaptic transmission is depressed in cultured
hippocampal neurons of APP/PS1 mice. Neurobiology of Aging, 26, 1227-37.
21. Kong, G.K.-., Adams, J.J. (2007) Structural Studies of the Alzheimer’s Amyloid Precursor
Protein Copper-binding Domain Reveal How it Binds Copper Ions, J Mol Biol, 367, 148.
22. Hasegawa, M., Arai, T. & Ihara, Y. (1990). Immunochemical evidence that fragments of
phosphorylated MAP5 (MAP1B) are bound to neurofibrillary tangles in Alzheimer's
disease. Neuron, 4, 909.
87
23. Perry, G., Friedman, R., Shaw, G. & Chau, V. (1987). Ubiquitin is detected in neurofibrillary
tangles and senile plaque neurites of Alzheimer disease brains. Proc Natl Acad Sci U S A,
84, 3033.
24. Rosenblatt, D.E., Geula, C. & Mesulam, M.M. (1989). Protease nexin I immunostaining in
Alzheimer's disease. Ann Neurol, 26, 5, 628.
25. Baum, L., Masliah, E., Iimoto, D.S., Hansen, L.A., Halliday, W.C. & Saitoh, T. (1992).
Casein kinase II is associated with neurofibrillary tangles but is not an intrinsic component
of paired helical filaments. Brain Res, 21, 126–132.
26. Wisniewski, K.E., Wisniewski, H.M. & Wen, G.Y. 1985, Occurrence of neuropathological
changes and dementia of Alzheimer's disease in Down's syndrome. Ann Neurol, 17, 278.
27. Turner, P.,R.. (2003). Roles of amyloid precursor protein and its fragments in regulating
neural activity, plasticity and memory, Prog Neurobiol, 70, 1-15.
28. Selkoe, D.J. (2001). Alzheimer's disease, genes, proteins, and therapy, Physiol Rev, 81, 741.
29. Findeis, M.A. (2007). The role of amyloid beta peptide 42 in Alzheimer's disease, Pharmacol
Ther, 116, 266.
30. Rogaeva, E., Meng, Y., Lee. (2007). The neuronal sortilin-related receptor SORL1 is
genetically associated with Alzheimer disease, Nat Genet, 39, 168.
31. Crouch P. J., Harding S. M. (2008). Mechanisms of Ab mediated neurodegeneration in
Alzheimer’s disease. Int. J. Biochem. Cell Biol. 40, 181– 198.
31. Hardy, J. & Selkoe, D.J. (2002). The Amyloid Hypothesis of Alzheimer's Disease, Progress
and Problems on the Road to Therapeutics, Science, 297, 353-356.
33. Gandy S, Simon AJ, (2010). Days to criterion as an indicator of toxicity associated with
human Alzheimer amyloid-beta oligomers. Ann Neurol 68, 220-230.
34. Feng, Y., Yang, S.G. (2009). Ellagic acid promotes A beta 42 fibrillization and inhibits A
beta 42-induced neurotoxicity, Biochem Biophys Res Commun, 390, 1250-1254.
88
35. Perry, V., Nicoll, J. & Holmes, C. (2010). Microglia in neurodegenerative disease, Nat Rev
Neurol, 6, 193-201.
36. H.W. Wang, J.F. Pasternak, H. (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.
37. B.L. Trommer, C. Shah. (2005). ApoE isoform-specific effects on LTP, blockade by
oligomeric amyloidbeta1–42, Neurobiol. Dis. 18, 75–82.
38. Yun, S.H., Gamkreidze, G., Stine, W.B. (2006). Amyloid-beta 1–42 reduces neuronal
excitability in mouse dentate gyrus, Neuroscience Letters, 403, 162-165.
39. Minkeviciene, R. (2009). Amyloid beta-induced neuronal hyperexcitability triggers
progressive epilepsy. J. Neurosci. 29, 3453–3462.
40. Palop, J.J., Mucke, L. (2010). Amyloid-β Induced Neuronal Dysfunction in Alzheimer’s
Disease, From Synapses toward Neural Networks, Nature Neuroscience, 13, 812-818.
43.32 Thal, D.,R., Rüb, U. (2002). Phases of A betadeposition in the human brain and its
relevance for the development of AD, Neurology, 58, 1791.
44. Holmes C. (2008). Long-term effects of Abeta42 immunisation in Alzheimer’s disease,
follow-up of a randomised, placebo-controlled phase I trial. Lancet, 372, 216-223.
45. MRC-CFAS Neuropathology Group. (2001). Pathological correlates of lateonset dementia in
a multicentre, community-based population in England and Wales. Neuropathology Group
of the Medical Research Council Cognitive.
46. Gómez-Isla, T., Hollister, R., West, H., Mui, S. (1997). Neuronal loss correlates with but
exceeds neurofibrillary tangles in Alzheimer's disease, Ann Neurol, 41, 17.
47. Iqbal, K. (2009). Mechanisms of tau-induced neurodegeneration, Acta Neuropathol, 118, 53-
69.
89
48. Lace, G.L., Wharton, S.B. & Ince, P.G.(2007). A brief history of tau, the evolving view of
the microtubule-associated protein tau in neurodegenerative diseases, Clin Neuropathol,
26,. 43.
49. Andreadis, A., Brown, W.M. & Kosik, K.S. (1992). Structure and novel exons of the human
tau gene, Biochemistry, 31 10626.
50. Duyckaerts, C., Delatour, B. & Potier, M. (2009). Classification and basic pathology of
Alzheimer disease, Acta Neuropathol, 118, 5-36
51. Heslenfeld D. (2003) Visual mismatch negativity. In, Polich J, editor. Detection of change,
event-related potential and fMRI findings. Boston, MA, Kluwer, 41–59.
52. Kujala A, Na¨a¨ta¨nen R. (2003). Auditory environment and change detection as indexed by
the mismarch gegativity (MMN). In, Polich J, editor. Detection of change, event-related
potential and fMRI findings. Boston, MA, Kluwer, 1–22.
53. Polich J. (1998). Habituation of P300 from auditory stimuli. Psychobiology, 17, 19–28.
54. Duncan-Johnson CC, Donchin E. (1977). Effects of a priori and sequential probability of
stimuli on event-related potential. Psychophysiology, 14-95.
55. Johnson R, Donchin E. (1982). Sequential expectancies and decision-making in a changing
environment, an electrophysiological approach. Psychophysiology, 19, 183–200.
56. Squires KC, Wickens C. (1976).. Squires NK, Donchin E. Effect of stimulus sequence on
waveform of cortical event-related potential. Science, 193, 1142–6.
57. Snyder E, Hillyard S. (1976). Long-latency evoked potentials to irrelevant, deviant stimuli.
Behav Biol, 16, 319–31.
58. Squires N, Squires K, Hillyard S. (1975). Two varieties of long-latency positive waves
evoked by unpredictable auditory stimuli in man. Electroencephalogr Clin Neurophysiol,
38, 387–401.
90
59. Courchesne E, Hillyard SA, Galambos R. (1975). Stimulus novelty, task relevance, and
visual evoked-potential in man. Electroencephalogr Clin Neurophysiol 39, 131–43.
60. Knight, R.T. (1984). Decreased response to novel stimuli after prefrontal lesions in man.
Electroencephalogr Clin Neurophysiol 59, 9–20.
61. Grunwald T, Lehnertz K, Heinze HJ, Helmstaedter C, Elger CE. (1998). Verbal novelty
detection within the hippocampus proper. Proc Natl Acad Sci USA, 95, 3193–7.
62. Knight RT. (1996). Contribution of human hippocampal region to novelty detection. Nature,
383, 256–9.
63. Bancher, C., Braak, H., Fischer, P., & Jellinger, K. a. (1993). Neuropathological staging of
Alzheimer lesions and intellectual status in Alzheimer’s and Parkinson's disease patients.
Neuroscience Letters, 162, 179–182.
64. Ferrer, I., Santpere, G., Leeuwen, F.W. (2008). Argyrophilic grain disease, Brain, 131, 1416-
1432.
65. Tolnay, M., Probst, A. (2003). The Neuropathological Spectrum of Neurodegenerative
Tauopathies, Life, 55(6), 399-305.
66 Rademakers, R. (2004). The role of tau (MAPT) in frontotemporal dementia and related
tauopathies, Human Mutation, 24, 277-95.
67. Thomas, P. & Fenech, M. (2007). A review of genome mutation and Alzheimer's disease,
Mutagenesis, 22. 15.
68. Martorana, A., Esposito, Z., & Koch, G., (2010). Beyond the cholinergic hypothesis, do
current drugs work in Alzheimer's disease?, CNS Neurosci Ther, 16, 235.
69. Contestabile, A. (2011). The history of the cholinergic hypothesis, Behav Brain Res, 221(2),
334-4.
70. Coyle, J.T., Price, D.L., Delong, M.R. (1983). Alzheimer's Disease, A Disorder of Cortical
Cholinergic Innervation, Science, 219, 1184-1190.
91
71. Bowen DM, Smith CB, White P, Davison AN. (1976) Neurotransmitter-related enzymes and
indices of hypoxia in senile dementia and other abiotrophies, Brain, 99, 459-69.
72. Davies, P., Maloney, A.J. (1976). Selective loss of central cholinergic neurons in Alzheimer's
disease. Lancet, 308, 1403.
73. Perry, E.K., Tomlinson, B.E., Blessed. G. (1978). Correlation of cholinergic abnormalities
with senile plaques and mental test scores in senile dementia, British Medical Journal, 2,
1457.
74. Davies, P. (1979). Neurotransmitter-related enzymes in senile dementia of the Alzheimer
type. Brain research, 171, 319-27.
75. Davies. P., Katzman, R., Terry, R.D. (1980). Reduced somatostatin-like immunoreactivity in
cerebral cortex from cases of Alzheimer disease and Alzheimer senile dementa, Nature,
288, 279-80.
76. Mufson E.J., Counts S.E., Fahnestock M, Ginsberg S.D. (2007). Cholinotrophic molecular
substrates of mild cognitive impairment in the elderly. Curr. Alzheimer Res 2007,4(4),
340–350.
77. Mesulam M, Shaw P, Mash D, Weintraub S.(2004). Cholinergic nucleus basalis tauopathy
emerges early in the aging-MCI-AD continuum. Ann. Neurol, 55(6), 815–828
78. Collerton, D. (1986). Cholinergic function and intellectual decline in Alzheimer’s disease,
Neuroscience, 19, 1-28
79. Bartus RT, Dean RL 3rd, Beer, B. (1982). The cholinergic hypothesis of geriatric memory
dysfunction, science, 217, 408-14.
80. Brinkman SD, Braun P.(1984). Classification of dementia patients by a WAIS profile related
to central cholinergic deficiencies. J Clin Neuropsychol., 6, 393-400.
81. Bartus RT, Dean RL 3rd, Beer, B. (1982). The cholinergic hypothesis of geriatric memory
dysfunction, science, 217, 408-14.
92
82. Brinkman SD, Braun P.(1984). Classification of dementia patients by a WAIS profile related
to central cholinergic deficiencies. J Clin Neuropsychol., 6, 393-400.
83. Longo, V.G. (1966). Behavioral and electroencephalographic effects of atropine and related
compounds, Pharmacological review, 18, 968-966.
84. 64 Drachman, D.A. (1977). Memory and cognitive function in man, does the cholinergic
system have a specific role? Neurology, 27, 783.
85. Davis, K.L., Mohs, R.C. (1978). Physostigmine, improvement of long-term memory
processes in normal humans, Science, 201, 272-4
86. Bierer LM, Haroutunian V, Gabriel S. (1995). Neurochemical correlates of dementia severity
in Alzheimer's disease, relative importance of the cholinergic deficits, Journal of
neurochemistry, 64, 749-60.
87. Gomez-Isla T, Price JL, McKeel DW Jr. (1996). Profound loss of layer II entorhinal cortex
neurons occurs in very mild Alzheimer's disease. J. Neurosci, 16(14), 4491–4500.
89. Hyman BT, Van Hoesen GW, Damasio AR, Barnes CL.(1984). Alzheimer's disease, cell-
specific pathology isolates the hippocampal formation. Science, 225, 1168–1170.
90. Kordower JH, Chu Y, Stebbins GT. (2001). Loss and atrophy of layer II entorhinal cortex
neurons in elderly people with mild cognitive impairment. Ann. Neurol, 49(2), 202–213.
91. Rogers SL, Doody RS, Pratt RD, Ieni JR (2000). Long-term efficacy and safety of donepezil
in the treatment of Alzheimer's disease, final analysis of a US multicentre open-label
study. Eur. Neuropsychopharmacol, 10, 195–203.
92. Martorana, A., Esposito, Z., & Koch, G. (2010). Beyond the cholinergic hypothesis, do
current drugs work in Alzheimer's disease?, CNS Neurosci Ther, 16, 235.
93. van der Flier, W.M., Pijnenburg, Y.A.L., Fox, N.C. & Scheltens, P. (2011). Early-onset
versus late-onset Alzheimer's disease, the case of the missing APOE epsilon 4 allele,
Lancet Neurol, 10, 280- 288.
93
94. Miyoshi, K. (2009). What is 'early onset dementia'?, Psychogeriatrics, 9, 2, 67.
95. Wisniewski, K.E., Wisniewski, H.M. & Wen, G.Y. (1985). Occurrence of neuropathological
changes and dementia of Alzheimer's disease in Down's syndrome, Ann Neurol, 17, 278.
96. Levy-Lahad, E., Wijsman, E.M. (1995). A familial Alzheimer's disease locus on
chromosome 1, Science, 269, 5226, 970.
97. St George-Hyslop, P., Haines, J.A. & Sorbi, S. (1992). Genetic evidence for a novel familial
Alzheimer's disease locus on chromosome 14, Nat Genet, 2, 4, 330.
99. Zhang, L., Song, L., Terracina, G., Liu, Y., Pramanik, B., Parker, E. (2001). Biochemical
characterization of the gamma-secretase activity that produces beta-amyloid peptides,
Biochemistry, 40, 16, 5049.
100. Kukull, W.A. & Ganguli, M. (2000). Epidemiology of dementia, concepts and overview,
Neurol Clin, 18, 4, 923.
101. Iqbal, K. & Grundke-Iqbal, I. (2010). Alzheimer's disease, a multifactorial disorder seeking
multitherapies, Alzheimers Dem, 6, 420.
102. Iqbal, K., Flory, M., Khatoon, S. (2005). Subgroups of Alzheimer's disease based on
cerebrospinal fluid molecular markers, Ann Neurol, 58, 5, 748.
103. Jellinger, K.A. & Attems, J. (2007). Neurofibrillary tangle-predominant dementia,
comparison with classical Alzheimer disease, Acta Neuropathol, 113, 2, 107.
104. Katzman, R., Terry, R., DeTeresa. (1988). Clinical, pathological, and neurochemical
changes in dementia, a subgroup with preserved mental status and numerous neocortical
plaques, Ann Neurol, 23, 2, 138.
105. Duyckaerts C, Delatour B, Potier M. (2009). Classification and basic pathology of
Alzheimer disease. Acta Neuropathol 118, 5-36.
106. Guy M. McKhann (2011). The diagnosis of dementia due to Alzheimer's disease,
Recommendations from the National Institute on Aging – Alzheimer's Association
94
workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & Dementia,
The Journal of the Alzheimer's Association, 7(3),263 – 269.
107. Marilyn S. Albert. (2011). The diagnosis of mild cognitive impairment due to Alzheimer's
disease, Recommendations from the National Institute on Aging – Alzheimer's
Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's &
Dementia, The Journal of the Alzheimer's Association, 7(3),270 – 279.
108. Reisa A. (2001). Toward defining the preclinical stages of Alzheimer's disease,
Recommendations from the National Institute on Aging – Alzheimer's Association
workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & Dementia,
The Journal of the Alzheimer's Association,7(3), 280–292.
109. Gold, C.A. (2008). Memory loss in Alzheimer's disease, implications for development of
therapeutics, Expert review of neurotherapeutics, 8, 1879-1891.
110. Blennow K, de Leon MJ, Zetterberg H. (2006). Alzheimer’s disease. Lancet. 368, 387–403
111. Goedert M, Spillantini MG. (2006). A century of Alzheimer’s disease. Science. 314, 777–
781.
112. Braak E, Griffing K, Arai K, Bohl J, Bratzke H, Braak H. (1999). Neuropathology of
Alzheimer’s disease, what is new since A. Alzheimer? Eur Arch Psychiatry Clin
Neurosci, 3, 14–22.
113. Schmitt FA, Davis DG, Wekstein DR, Smith CD. (2000). Preclinical AD revisited,
neuropathology of cognitively normal older adults. Neurology. 55 (3), 370–376.
114. Giannakopoulos P, Hof PR, Mottier S, Michel JP, Bouras C. (1994). Neuropathological
changes in the cerebral cortex of 1258 cases from a geriatric hospital, retrospective
clinicopathological evaluation of a 10-year autopsy population. Acta Neuropathol., 87
(5), 456–468.
115. Guillozet, A.L, Weintraub, S. (2003). Neurofibrillary tangles, amyloid, and memory in
aging and mild cognitive impairment. Arch Neurol., 60 (5), 729–736.
95
116. Tomlinson BE, Blessed G, Roth M. (1968). Observations on the brains of non-demented old
people. J Neurol Sci., 7 (2), 331–356.
117. Morris JC, Price AL.(2001). Pathologic correlates of nondemented aging, mild cognitive
impairment, and early-stage Alzheimer’s disease. J Mol Neurosci, 17 (2), 101–118.
118. Thal DR, Rub U, Schultz C, Sassin I, Ghebremedhin E, Del Tredici K, Braak E, Braak H.
(2000). Sequence of Abeta-protein deposition in the human medial temporal lobe. J
Neuropathol Exp Neurol, 59 (8), 733–748.
119. Dubois B, Feldman HH, Jacova C. (2007). Research criteria for the diagnosis of
Alzheimer’s disease, revising the NINCDS-ADRDA criteria. Lancet Neurol. 6 (8), 734–
746.
120. Price JL, Ko AI, Wade MJ, Tsou SK, McKeel DW, Morris JC (2001). Neuron number in
the entorhinal cortex and CA1 in preclinical Alzheimer disease. Arch Neurol. 58 (9),
1395–1402.
121. Katzman R, Terry R, DeTeresa R, Brown T, Davies P, Fuld P, Renbing X, Peck A. (1988).
Clinical, pathological, and neurochemical changes in dementia, a subgroup with
preserved mental status and numerous neocortical plaques. Ann Neurol. 23 (2), 138–144.
122. Gomez-Isla T, Price JL, McKeel DW, Jr., Morris JC (1996). Profound loss of layer II
entorhinal cortex neurons occurs in very mild Alzheimer’s disease. J Neurosci. 16 (14),
4491–4500.
123. West MJ, Coleman PD, Flood DG, Troncoso JC.( 1994). Differences in the pattern of
hippocampal neuronal loss in normal ageing and Alzheimer’s disease. Lancet., 344,
769–772.
124. West MJ, Kawas CH, Stewart WF, Rudow GL, Troncoso JC. (2004). Hippocampal neurons
in pre-clinical Alzheimer’s disease. Neurobiol Aging, 25 (9), 1205– 1212.
125. Kordower JH, Chu Y, Stebbins GT, DeKosky ST, Cochran EJ, Bennett D, Mufson EJ.
(2001). Loss and atrophy of layer II entorhinal cortex neurons in elderly people with
mild cognitive impairment. Ann Neurol, 49 (2), 202–213.
96
126. Price JL. (2003). Aging, preclinical Alzheimer disease, and early detection. Alzheimer Dis
Assoc Disord., 17, 60–62.
127. Coleman PD, Flood DG. (1987). Neuron numbers and dendritic extent in normal aging and
Alzheimer’s disease. Neurobiol Aging. 8 (6), 521–545.
128. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders.
Fourth Edition (DSM-IV). Washington D.C., American Psychiatric Association, 1994.
129. World Health Organisation. The ICD-10. International Classification of Mental and
Behavioural Disorders, Clinical Descriptions and Diagnostic Guidelines. Geneva,
Switzerland, World Health Organisation, 1992.
130. Forstl H, Kurz A. (1999). Clinical features of Alzheimer’s disease. Eur Arch Psychiatry
Clin Neurosci, 249 (6), 288–290.
131. Nestor PJ, Scheltens P, Hodges JR. (2004). Advances in the early detection of Alzheimer’s
disease. Nat Med., 10, 34–41.
132. Salmon DP, Bondi MW. (2009). Neuropsychological assessment of dementia. Annu Rev
Psychol. 60, 257–282.
133. Almkvist O, Backman L. (1993). Detection and staging of early clinical dementia. Acta
Neurol Scand. 88 (1), 10–15.
134. Almkvist O, Winblad B. (1999). Early diagnosis of Alzheimer dementia based on clinical
and biological factors. Eur Arch Psychiatry Clin Neurosci. 249 Suppl 3, 3–9.
135. Backman L, Wahlin A, Small BJ, Herlitz A, Winblad B, Fratiglioni L. (2004). Cognitive
Functioning in Aging and Dementia, The Kungsholmen Project. Aging,
Neuropsychology, and Cognition, 11 (2), 212–244.
136. Perry RJ, Hodges JR. (1999). Attention and executive deficits in Alzheimer’s disease. A
critical review. Brain, 122, 383–404.
97
137. Perry RJ, Hodges JR. (2000) Alzheimer’s disease in memory impaired patients (very mild)
Alzheimer’s disease, longitudinal profiles of individual subjects’ decline. Dement
Geriatr Cogn Disord., 11 (6), 342–349.
138. Crystal H, Dickson D. (1988). Clinico-pathologic studies in dementia, nondemented
subjects with pathologically confirmed Alzheimer’s disease. Neurology. 38 (11), 1682–
1687.
139. Galvin JE, Powlishta KK, Wilkins K, McKeel DW, Jr., Xiong C, Grant E, Storandt M,
Morris JC. (2005). Predictors of preclinical Alzheimer disease and dementia, a
clinicopathologic study. Arch Neurol. 62 (5), 758–765.
140. Markesbery WR, Schmitt FA, Kryscio RJ, Davis DG, Smith CD, Wekstein DR. (2006).
Neuropathologic substrate of mild cognitive impairment. Arch Neurol. 63 (1), 38–46.
141. Storandt M, Grant EA, Miller JP, Morris JC. (2006). Longitudinal course and
neuropathologic outcomes in original vs revised MCI and in pre-MCI. Neurology. 67
(3), 467–473.
142. Morris JC, Storandt M, Miller JP, McKeel DW, Price JL, Rubin EH, Berg L. (2001). Mild
cognitive impairment represents early-stage Alzheimer disease. Arch Neurol, 58 (3)
397–405.
143. Gauthier S, Reisberg B, Zaudig M, Petersen RC. (2006). Mild cognitive impairment.
Lancet, 367, 1262–1270.
144. Tierney MC, Szalai JP, Snow WG. (1996). Prediction of probable Alzheimer’s disease in
memory impaired patients, A prospective longitudinal study. Neurology, 46 (3), 661–
665.
145. Estevez-Gonzalez A, Kulisevsky J, Boltes A, Otermin P, Garcia-Sanchez C. (2003). Rey
verbal learning test is a useful tool for differential diagnosis in the preclinical phase of
Alzheimer’s disease, comparison with mild cognitive impairment and normal aging. Int
J Geriatr Psychiatry. 18 (11), 1021–1028
98
146 Albert MS, Moss MB, Tanzi R, Jones K. (2001). Preclinical prediction of AD using
neuropsychological tests. J Int Neuropsychol Soc., 7 (5), 631–639.
147. Backman L, Small BJ, Fratiglioni L. (2001). Stability of the preclinical episodic memory
deficit in Alzheimer’s disease. Brain, 96–102.
148. Arnaiz E, Almkvist O. ( 2003). Neuropsychological features of mild cognitive impairment
and preclinical Alzheimer’s disease. Acta Neurol Scand Suppl, 179, 34–41.
149. Backman L, Jones S, Berger AK, Laukka EJ, Small BJ. (2005). Cognitive impairment in
preclinical Alzheimer’s disease, a meta-analysis. Neuropsychology, 19 (4), 520–531
150. Arnaiz E, Almkvist o. (2004). Mild cognitive impairment, a cross-national comparison. J
Neurol Neurosurg Psychiatry. 75 (9), 1275–1280.
151. Sacuiu S, Sjogren M, Johansson B, Gustafson D, Skoog I. (2005). Prodromal cognitive
signs of dementia in 85-year-olds using four sources of information. Neurology. 65 (12),
1894–1900.
152. Bennett DA, Schneider JA, Arvanitakis Z, Kelly JF, Aggarwal NT, Shah RC, Wilson RS.
(2006). Neuropathology of older persons without cognitive impairment from two
community-based studies. Neurology. 66 (12), 1837–1844.
153. Hulette CM, Welsh-Bohmer KA. (1998). Neuropathological and neuropsychological
changes in normal aging, evidence for preclinical Alzheimer disease in cognitively
normal individuals. J Neuropathol Exp Neurol. 57 (12), 1168–1174.
154. Price JL, McKeel DW, Jr., Buckles VD. (2009). Neuropathology of nondemented aging,
presumptive evidence for preclinical Alzheimer disease. Neurobiol Aging. 30 (7), 1026–
1036.
155. Mackenzie IR. (1994). Senile plaques do not progressively accumulate with normal aging.
Acta Neuropathol. 87 (5), 520–525
99
156. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. (2007). Mixed brain pathologies
account for most dementia cases in community-dwelling older persons. Neurology. 69
(24), 2197–2204.
157. Price JL, Davis PB, Morris JC, White DL. (1991). The distribution of tangles, plaques and
related immunohistochemical markers in healthy aging and Alzheimer’s disease.
Neurobiol Aging. 12 (4), 295–312.
158. Knopman DS, Parisi JE, Petersen LE. (2003). Neuropathology of cognitively normal
elderly. J Neuropathol Exp Neurol, 62 (11), 1087–1095.
159. Goldman WP, Price JL, Storandt M, Grant EA, McKeel DW, Jr., Rubin EH, Morris JC.
(2001). Absence of cognitive impairment or decline in preclinical Alzheimer’s disease.
Neurology. 56 (3), 361–367.
160. Neuropathology Group of the Medical Research Council Cognitive Function and Ageing
Study (MRC CFAS). Pathological correlates of late-onset dementia in a multicentre,
community-based population in England and Wales. (2001). Lancet, 357 169–175.
161. Hansen LA, DeTeresa R, Davies P, Terry RD. (1988). Neocortical morphometry, lesion
counts, and choline acetyltransferase levels in the age spectrum of Alzheimer’s disease.
Neurology. 38 (1), 48–54.
162. Jack, C.R. (2010). Hypothetical model of dynamic biomarkers of the Alzheimer's
pathological cascade. Lancet Neurol. 9, 119-128.
163. Post, S. G. (1999). Future scenarios for the prevention and delay of Alzheimer disease onset
in high-risk groups, An ethical perspective. American Journal of Preventive Medicine,
16, 105–110.
164. Elias MF, Beiser A, Wolf PA, Au R, White RF, D’Agostino RB. (2000). The preclinical
phase of alzheimer disease, A 22-year prospective study of the Framingham Cohort.
Arch Neurol, 57 (6), 808–813.
100
165. Palmer K, Backman L, Winblad B, Fratiglioni L. (2008). Early symptoms and signs of
cognitive deficits might not always be detectable in persons who develop Alzheimer’s
disease. Int Psychogeriatr, 20 (2), 252–258.
166 Small BJ, Fratiglioni L, Viitanen M, Winblad B, Backman L. The course of cognitive
impairment in preclinical Alzheimer disease, three- and 6-year followup of a population-
based sample. Arch Neurol., 57 (6), 839–844.
167. Tierney MC, Szalai JP, Dunn E, Geslani D, McDowell I. (2000). Prediction of probable
Alzheimer disease in patients with symptoms suggestive of memory impairment. Value
of the Mini-Mental State Examination. Arch Fam Med., 9 (6), 527– 532
169. Masur DM, Sliwinski M, Lipton RB, Blau AD, Crystal HA.( 1994). Neuropsychological
prediction of dementia and the absence of dementia in healthy elderly persons.
Neurology., 44 (8), 1427–1432.
170. Tierney MC, Yao C, Kiss A, McDowell I. (2005). Neuropsychological tests accurately
predict incident Alzheimer disease after 5 and 10 years. Neurology., 64 (11), 1853–
1859.
171. Twamley EW, Ropacki SA, Bondi MW. (2006). Neuropsychological and neuroimaging
changes in preclinical Alzheimer’s disease. J Int Neuropsychol Soc., 12 (5), 707–735.
172. Small, B. J., Herlitz, A., Fratiglioni, L., Almkvist, O., & Ba¨ckman, L. (1997). Cognitive
predictors of incident Alzheimer’s disease, A prospective longitudinal study.
Neuropsychology, 11, 413–420.
173. Grober, E., & Kawas, C. (1997). Learning and retention in preclinial and early Alzheimer’s
disease. Psychology and Aging, 12, 183–188.
174. Fox, N. C., Warrington, E. K., Freeborough, P. A., Hartikainen, P., Kennedy, A. M.,
Stevens, J. M., & Rossor, M. N. (1996). Presymptomatic hippocampal atrophy in
Alzheimer’s disease. Brain, 119, 2001– 2007.
101
175. Ba¨ckman, L., Small, B. J., & Wahlin, Å. (2001). Aging and memory, Cognitive and
biological perspectives. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the
psychology of aging (5th ed., 349–377). San Diego, Academic Press.
176. Park, D. C., Lautenschalger, G., Hedden, T., Davidson, N. S., Smith, A. D., & Smith, P. K.
(2002). Models of visuospatial and verbal memory across the adult life span. Psychology
and Aging, 17, 299–320.
177. Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition.
Psychological Review, 103, 403–428.
178. Launer, L. J., Ross, G. W., Petrovitch, H., Masaki, K. H., Foley, D. J., White, L. R., &
Havlik, R. J. (2000). Midlife blood pressure and dementia, The Honolulu-Asia aging
study. Neurobiology of Aging, 21, 49–55.
179. Plassman, B. L., Havlik, R. J., Steffens, D. C., Helms, M. J., Newman, T. N., Drosdick, D.,
et al. (2000). Documented head injury in early adulthood and risk of Alzheimer’s
disease and other dementias. Neurology, 55, 1158–1166.
180. Snowdon, D. A., Greiner, L. H., Mortimer, J. A., Riley, K. P., Greiner, P. A., &
Markesbery, W. R. (1997). Brain infarction and the clinical expression of Alzheimer’s
disease, The Nun Study. Journal of the American Medical Association, 277, 813–817.
181. Wilson, R. S., Evans, D. A., Bienias, J. L., de Leon, C. F. M., Schneider, J. A., & Bennett,
D. A. (2003). Proneness to psychological distress is associated with risk of Alzheimer’s
disease. Neurology, 61, 1479–1485.
182. .Fox, N. C., Crum, W. R., Scahill, R. I., Stevens, J. M., Janssen, J. C., & Rossor, M. N.
(2001). Imaging of onset and progression of Alzheimer’s disease with voxel-
compression mapping of serial magnetic images. Lancet, 358, 201–205
183. .Killiany, R. J., Gomez-Isla, T., Moss, M., Kikinis, R., Sandor, T., Jolesz, F., et al. (2000).
Use of structural magnetic resonance imaging to predictwho will get Alzheimer’s
disease. Annals of Neurology, 47, 430–439.
102
184. van der Flier, W. M., van den Heuvel, D. M. J., Weverling-Rijnsburger, A. W. E., Bollen,
E. L. E. M., Westendorp, G. J., van Buchem, M. A.,& Middlekoop, H. A. M. (2002).
Magnetization transfer imaging in normal aging, mild cognitive impairment, and
Alzheimer’s disease. Annals of Neurology, 52, 62–67.
185. Yamaguchi, H., Sugihara, S., Ogawa, A., Oshima, N., & Ihara, Y. (2001). Alzheimer beta
amyloid deposition enhanced by ApoE epsilon 4 geneprecedes neurofibrillary pathology
in the frontal association cortex of nondemented senior subjects. Journal of
Neuropathology and Experimental Neurology, 60, 731–739
186. Brier, M.R., Gordon, B., Friedrichsen, K. (2016). Tau and Ab imaging, CSF measures, and
cognition in Alzheimer’s disease, Alzheimer’s Disease, 8, 338-66
187. Arna´iz, E., Jelic, V., Almkvist, O., Wahlund, L. O., Winblad, B., Valind, S., & Nordberg,
A. (2001). Impaired cerebral glucose metabolism andcognitive functioning predict
deterioration in mild cognitive impairment. NeuroReport, 12, 851–855.
188. Silverman, D. H. S., Small, G. W. (2001). Positron emission tomography in evaluation of
dementia, Regional brain metabolism and long-term outcome. Journal of the American
Medical Association, 286, 2120–212.
189. Kogure, D., Matsuda, H., Ohnishi, T. (2000). Longitudinal evaluation of early Alzheimer’s
disease using brain perfusion SPECT. Journal of Nuclear Medicine, 41, 1155–1162.
190. Small, S. A., Perara, G., DeLaPaz, R., Mayeux, R., & Stern, Y. (1999). Differential regional
dysfunction of the hippocampal formation among elderly with memory decline and
Alzheimer’s disease. Annals of Neurology, 45, 466–472
191. Geerlings, M. I., Jonker, C., Bouter, L. M., Ade.r, H. J., & Schmand, B. (1999). Association
between memory complaints and incident Alzheimer’s disease in elderly people with
normal baseline cognition. American Journal of Psychiatry, 156, 531–537
192. Palmer, K., Ba¨ckman, L., Winblad, B., & Fratiglioni, L. (2003). Detection of Alzheimer’s
disease and dementia in the preclinical phase, Predictivity of a three-step procedure in a
population-based study. British Medical Journal, 326, 245–247
103
193. Berger, A-K., Fratiglioni, L., Forsell, Y., Winblad, B., & Ba¨ckman, L. (1999). The
occurrence of depressive symptoms in the preclinical phase of AD, A population-based
study. Neurology, 53, 1998–2002.
194. Fox N, Growdon JH. (2004). Biomarkers and Surrogates. NeuroRx, the journal of the
American Society for Experimental NeuroTherapeutics, 1 (2), 181.
195. Sunderland T, Hampel H, Takeda M, Putnam KT, Cohen RM. (2006). Biomarkers in the
diagnosis of Alzheimer’s disease, are we ready? J Geriatr Psychiatry Neurol., 19 (3),
172–179.
196. SBU (2006). Dementia diseases. Vol 172. First ed. Stockholm, SBU, The Swedish Council
on Technology Assessment in Health Care.
197. Saunders AM, Strittmatter WJ, Schmechel D. (1993). Association of apolipoprotein E allele
e4 with late-onset familial and sporadic Alzheimer’s disease. Neurology, 43, 1467-1472.
199. Gomez-Isla T, West HL, Rebeck GW. (1996). Clinical and pathological correlates of
apolipoprotein E e4 in Alzheimer’s disease. Ann Neurol.,39,62-70.
200. Mayeux R, Saunders AM, Shea S. (1998). Utility of the apolipoprotein E genotype in the
diagnosis of Alzheimer’s disease. N Engl J Med.,338,506-511.
200. Liu, C.C. (2013). Apolipoprotein E and Alzheimer disease, risk, mechanisms and therapy.
Nature Reviews. Neurology. 9 (2), 106–18.
201. Tur-Perez, J., Campion, D. (1995). Evidence for Apolipoprotein E €4 Association in Early-
Onset Alzheimer’s Patients With Late-Onset Relatives, American Journal of Medical
Genetics, 60, 550-553
202. Saunders, A.M., Strittmatter, W.J. (1993). Association of apolipoprotein E allele €4 with
late-onset familial and sporadic Alzheimer’s disease. Neurology 13,1467-1472.
203. Mayeux R, Stem Y, Ottman R, Tatemichi TK, Tang M-X, Maestre G, Ngai C, Tycko B,
Ginsberg H (1993). The apolipoprotein €4 allele in patients with Alzheimer’s disease.
Ann Neurol, 34,752-754.
104
204. Corder, E.H., Saunders, A.M. (1993). Gene dose of apolipoprotein E type 4 allele and the
risk ofAlzheimer’s disease in late onset families. Science 261,921-923.
205. 175 Growdon, J.H. (1999). Biomarkers of Alzheimer Disease, Arch Neuroogy, 56, 281-283
206. Palmer, K., Ba¨ckman, L., Winblad, B., & Fratiglioni, L. (2003). Detection of Alzheimer’s
disease and dementia in the preclinical phase, Predictivity of a three-step procedure in a
population-based study. British Medical Journal, 326, 245–247
207. Berger, A-K., Fratiglioni, L., Forsell, Y., Winblad, B., & Ba¨ckman, L. (1999). The
occurrence of depressive symptoms in the preclinical phase of AD, A population-based
study. Neurology, 53, 1998–2002.
208. Tapiola T, Alafuzoff I, Herukka SK. (2009). Cerebrospinal fluid {beta}-amyloid 42 and tau
proteins as biomarkers of Alzheimer-type pathologic changes in the brain. Archives of
Neurology, 66, 382–9.
209. Blennow K, Blennow K. (2004). Cerebrospinal fluid protein biomarkers for Alzheimer’s
disease. NeuroRx, 1, 213–25.
210. Zhang Y, Londos E, Minthon L. (2009). Medial temporal lobe atrophy increases the
specificity of cerebrospinal fluid biomarkers in Alzheimer disease with minor
cerebrovascular changes. Acta Radiologica, 50, 674–81.
211. Vemuri P, Wiste HJ, Weigand SD. (2009). MRI and CSF biomarkers in normal, MCI, and
AD subjects, diagnostic discrimination and cognitive correlations. Neurology, 73,287–
93.
212. Diniz, B.S, Pinto Junior, J.A. (2008) Do CSF total tau, phosphorylated tau, and beta-
amyloid 42 help to predict progression of mild cognitive impairment to Alzheimer’s
disease? A systematic review and meta-analysis of the literature. World Journal of
Biological Psychiatry ,9,172–82.
213. Mattsson N, Zetterberg H, Hansson O. (2009). CSF biomarkers and incipient Alzheimer
disease in patients with mild cognitive impairment. Journal of American Medical
Association, 302,385–93
105
214. Brys M, Pirraglia E, Rich K, (2009). Prediction and longitudinal study of CSF biomarkers
in mild cognitive impairment. Neurobiology of Aging,30,682–90.
215. Shaw, L.M, Vanderstichele, H, Knapik-Czajka M. (2009) .Cerebrospinal fluid biomarker
signature in Alzheimer’s disease neuroimaging initiative subjects. Annals of Neurology
,65,403–13.
216. Snider BJ, Fagan AM, Roe C. (2009). Cerebrospinal fluid biomarkers and rate of cognitive
decline in very mild dementia of the Alzheimer type. Archives of Neurology, 66,638–45.
217. 185 Peskind ER, Riekse R, Quinn JF. (2005). Safety and acceptability of the research
lumbar puncture. Alzheimer Disease & Associated Disorders, 19, 220–5.
218. Schneider, P., Hampel, H., Buerger, K. (2009). Biological marker candidates of
Alzheimer‘s disease in blood, plasma, and serum. CNS Neuroscience & Therapeutics,
15, 358–374.
219. Roher AE, Esh CL, Kokjohn TA, (2009). Amyloid beta peptides in human plasma and
tissues and their significance for Alzheimer’s disease. Alzheimer’s & Dementia, 5,18–
29.
220. Sutphen, C.L., Fagan, A.M. (2014). Progress Update, Fluid and Imaging Biomarkers in
Alzheimer’s Disease, Biol Psychiatry, 75, 520-526.
221. Aβ42Mayeux, R., Schupf, N. (2011). Blood-based biomarkers for Alzheimer’s disease,
Plasma Aβ40 and Aβ42, and genetic variants. Neurobiol Aging, 32 (suppl 1),S10–S19.
222. Fagan AM, Mintun MA, Shah AR. (2009), Cerebrospinal fluid tau and ptau181 increase
with cortical amyloid deposition in cognitively normal individuals, Implications for
future clinical trials of Alzheimer’s disease. EMBO Mol Med, 1, 371–380.
223. Ingelson M, Blomberg M, Benedikz E. (1999), Tau immunoreactivity detected in human
plasma, but no obvious increase in dementia. Dement Geriatr Cogn, 10,442–445.
224. Zetterberg H, Wilson D, Andreasson U, Minthon L, Blennow K.(2013). Plasma tau levels in
Alzheimer’s disease. Alzheimers Res Ther 5, 9.
106
225. O’Bryant SE, Xiao G, Barber R, Huebinger R, Wilhelmsen K, Edwards M. (2011), A
blood-based screening tool for Alzheimer’s disease that spans serum and plasma,
Findings from TARC and ADNI. PLoS One 6,e28092.
226. Reddy MM, Wilson R, Wilson J, Connell S, Gocke A, Hynan L.. (2011). Identification of
candidate IgG biomarkers for Alzheimer’s disease via combinatorial library screening.
Cell, 144, 132–142
227. Ray S, Britschgi M, Herbert C. (2007). Classification and prediction of clinical Alzheimer’s
diagnosis based on plasma signaling proteins. Nature Medicine, 13,1359–62.
228. Vemuri P, Wiste HJ, Weigand SD. (2009). MRI and CSF biomarkers in normal, MCI, and
AD subjects, diagnostic discrimination and cognitive correlations. Neurology,73, 287–
93.
229. Hua X, Leow AD, Parikshak N. (2008) Tensor-based morphometry as a neuroimaging
biomarker for Alzheimer’s disease, an MRI study of 676 AD, MCI, and normal subjects.
Neuroimage, 43,458–69.
230. Barnes J, Bartlett JW, van de Pol LA, (2009). A meta-analysis of hippocampal atrophy rates
in Alzheimer’s disease. Neurobiology of Aging, 30,1711–23.
231. Allan, C.L., Sexton, C.E. (2010). Imaging and biomarkers for Alzheimer’s disease,
Maturitas, 65, 138-142.
232. Vemuri P, Whitwell JL, Kantarci K. (2008). Antemortem MRI based Structural
Abnormality iNDex (STAND)-scores correlate with postmortem Braak neurofibrillary
tangle stage. Neuroimage, 42, 559–67.
233. Liu Y, Paajanen T, Zhang Y. (2009). Combination analysis of neuropsychological tests and
structural MRI measures in differentiating AD, MCI and control groups—The
AddNeuroMed study. Neurobiology of Aging.
234. Hua X, Leow AD, Parikshak N (2008). Tensor-based morphometry as a neuroimaging
biomarker for Alzheimer’s disease, an MRI study of 676 AD, MCI, and normal subjects.
Neuroimage, 43,458–69.
107
235. Devanand DP, Liu X, TabertMH. (2008). Combining early markers strongly predicts
conversion from mild cognitive impairment to Alzheimer’s disease. Biological
Psychiatry, 64,871–9.
236. Ebmeier KP, Filippini N, Mackay CE. Brain function and connectivity in dementia. In,
Ames D, Burns A, O’Brien JT, editors. Dementia. London, Hodder & Stoughton, in
press.
237. Gemmell HG, Sharp PF, Besson J. (1987). Differential-diagnosis in dementia using the
cerebral blood-flow agent Tc-99m Hmpao—a SPECT Study. Journal of Computer
Assisted Tomography,11, 398–402.
238. Raichle ME, Mintun MA. (2006). Brain work and brain imaging. Annual Review of
Neurosciences, 29, 449–76.
239 . Roy CS, Sherrington CS. (1890). On the regulation of the blood-supply of the brain.
Journal of Physiology, 11(85-158),117.
240. Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A. (2001). Neurophysiological
investigation of the basis of the fMRI signal. Nature, 412, 150–7.
241. Minoshima, S., Giordani, B., (1997). Metabolic reduction in the posterior cingulate cortex
in very early Alzheimer’s disease. Annals of Neurology, 42,85–94.
242. Rossor MN, Kennedy AM, Frackowiak RSJ. (1996). Clinical and neuroimaging featuresof
familial Alzheimer’s disease. Neurobiology of Alzheimer’s Disease, 777, 49–56.
243. Elgh E, Sundstrom T, Nasman B, Ahlstrom R, Nyberg L. (2002). Memory functions
andrCBF (99m)Tc-HMPAO SPET, developing diagnostics in Alzheimer’s disease
.European Journal of Nuclear Medicine and Molecular Imaging, 29,1140–8.
244. Ishii K, Sakamoto S. (1998). Cerebral glucose metabolism in patientswith frontotemporal
dementia. Journal of Nuclear Medicine, 39, 1875–8.
108
245. Lee YC, Liu RS, Liao YC. (2003). Statistical parametric mapping of brain SPECT
perfusionabnormalities in patients with Alzheimer’s disease. European Neurology, 49,
142–5
246. Villa G,Cappa A, Tavolozza M. (1995). Neuropsychological tests and [99mTc]-HM PAO
SPECT in the diagnosis of Alzheimer’s dementia. Journal of Neurology, 242, 359–66.
247. Holman BL, Johnson KA, Gerada B, Carvalho PA, Satlin A. (1992). The
scintigraphicappearance of Alzheimers-disease—a prospective-study using technetium-
99m-HMPAO SPECT. Journal of Nuclear Medicine, 33,181–5.
248. Alexander GE, Chen K, Pietrini P, Rapoport SI, Reiman EM. (2002). Longitudinal
PETevaluation of cerebral metabolic decline in dementia, a potential outcome measurein
Alzheimer’s disease treatment studies. American Journal of Psychiatry, 159,738–45
249. Herholz K, Salmon E, Perani D (2002). Discrimination between Alzheimer dementiaand
controls by automated analysis of multicenter FDG PET. Neuroimaging, 17, 302–16.
250. Hoffman JM, Welsh-Bohmer KA, Hanson M. (2000). FDG PET imaging inpatients with
pathologically verified dementia. Journal of Nuclear Medicine, 41, 1920–8.
251. Okamura N, Arai H, Higuchi M. (2001). [18F]FDG-PET study in dementia withLewy
bodies and Alzheimer’s disease. Progress in Neuropsychopharmacology and Biological
Psychiatry, 25 ,447–56.
252. Volkow ND, Zhu W, Felder, CA. (2002). Changes in brain functional homogeneityin
subjects with Alzheimer’s disease. Psychiatry Research Neuroimaging, 114, 39–50.
253. Masterman DL, Mendez MF, Fairbanks LA, Cummings JL.(1997). Sensitivity,
specificity,and positive predictive value of technetium99-HMPAOSPECT in
discriminating Alzheimer’s disease from other dementias. Journal of Geriatric
Psychiatry and Neurology, 10, 15–21.
254. Borroni B, Anchisi D, Paghera B. (2006). Combined 99mTc-ECD SPECT and
neuropsychological studies in MCI for the assessment of conversion to AD.
Neurobiology of Aging, 27,24–31.
109
255. Nobili F, De Carli F, Frisoni GB, (2009). SPECT predictors of cognitive decline and
Alzheimer’s disease in mild cognitive impairment. Journal of Alzheimers Diseases, 17,
761-72.
256. Staffen W, Bergmann J, Schonauer U. (2009). Cerebral perfusion (HMPAO-SPECT) in
patients with depression with cognitive impairment versus those with mild cognitive
impairment and dementia of Alzheimer’s type, a semiquantitative and automated
evaluation. European Journal of Nuclear Medicine and Molecular Imaging,36,801–10.
257. Habert MO, Horn JF, Sarazin M. (2009) Brain perfusion SPECT with an automated
quantitative tool can identify prodromal Alzheimer’s disease among patients with mild
cognitive impairment. Neurobiology of Aging, 32, 15-23.
258. Wolf H, Jelic V, Gertz HJ, Nordberg A, Julin P, Wahlund LO. (2003) A critical discussion
of the role of neuroimaging in mild cognitive impairment. Acta Neurologica
Scandinavica, Supplement,179,52–76.
259. De Santi S, de Leon MJ, Rusinek H. (2001) Hippocampal formation glucose metabolism
and volume losses in MCI and AD. Neurobiology of Aging, 22,529–39.
260. Jagust W, Thisted R, Devous Sr MD. (2001) SPECT perfusion imaging in the diagnosis of
Alzheimer’s disease, a clinical-pathologic study. Neurology, 56,950–6.
261. Dougall NJ, Bruggink S, Ebmeier KP. Clinical use of SPECT in dementia—a quantitative
review. In, Ebmeier KP, editor. SPECT in Dementia. Basel, Karger Verlag, 2003. p. 4–
37.
262. Ebmeier KP. Is there still a place for perfusion SPECT in the diagnosis of dementia? The
Open Nuclear Medicine Journal, in press.
263. Herholz K, Nordberg A, Salmon E. (1999). Impairment of neocortical metabolism predicts
progression in Alzheimer’s disease. Dementia and Geriatric Cognitive Disorders,
10,494–504.
264. Huang C, Eidelberg D, Habeck C. (2007). Imaging markers of mild cognitive impairment,
multivariate analysis of CBF SPECT. Neurobiology of Aging, 28, 1062–9.
110
265. Barrio JR, Huang SC, Cole G. (1999). Pet imaging of tangles and plaques in Alzheimer
disease with a highly hydrophobic profile. Journal of Labelled Compounds and
Radiopharmaceuticals, 42.
266. Mathis CA, Wang Y, Holt DP, Huang GF, Debnath ML, Klunk WE. (2003) Synthesis and
evaluation of 11C-labeled 6-substituted 2-arylbenzothiazoles as amyloid imaging agents.
Journal of Medicinal Chemistry, 46, 2740–54.
267. Agdeppa ED, Kepe V, Liu J. (2001). Binding characteristics of radiofluorinated 6-
dialkylamino-2-naphthylethylidene derivatives as positron emission tomography
imaging probes for beta-amyloid plaques in Alzheimer’s disease. Journal of
Neurosciences,21,RC189.
268. Small GW, Kepe V, Ercoli LM (2006) PET of brain amyloid and tau in mild cognitive
impairment. New England Journal of Medicine, 355, 2652–63.
269. Shoghi-Jadid K, Small GW, Agdeppa ED (2002) Localization of neurofibrillary tangles and
beta-amyloid plaques in the brains of living patients with Alzheimer disease. American
Journal of Geriatric Psychiatry, 10, 24–35.
270. Ikonomovic MD, Klunk WE, Abrahamson E. (2008). Post-mortem correlates of in vivo
PiB-PET amyloid imaging in a typical case of Alzheimer’s disease. Brain, 131,1630–45.
271. Klunk WE, Engler H, Nordberg A, (2004) Imaging brain amyloid in Alzheimer’s disease
with Pittsburgh Compound-B. Annals of Neurology, 55, 306–19.
272. Rowe CC, Ng S, Ackermann U. (2007) Imaging beta-amyloid burden in aging and
dementia. Neurology, 68, 1718–25
273. Forsberg A, Engler H, Almkvist O. (2008) PET imaging of amyloid deposition in patients
with mild cognitive impairment. Neurobiology of Aging, 29, 1456–65.
274. Ho, L., Fivecoat, H., Wang, J., Pasinetti, G.M. (2010). Alzheimer‘s disease biomarker
discovery in symptomatic and asymptomatic patients, experimental approaches and
future clinical applications. Experimental Gerontology, 45, 15–22.
111
275. Klunk, W.E., Mathis, C.A. (2008). The future of amyloid-beta imaging, a tale of
radionuclides and tracer proliferation. Current Opinion in Neurology, 21, 683–687.
276. Gordon, B., Friedrichsen, K., Brier, M. (2016). The relationship between cerebrospinal fluid
markers of Alzheimer pathology and positron emission tomography tau imaging, Brain,
138, 1-12.
277. Tatum, William O. (2014). Ellen R. Grass Lecture, Extraordinary EEG. Neurodiagnostic
Journal, 54.1, 3–21.
278. Dauwels, J., Vialatter, F., Cichrocki, A. (2010). Diagnosis of Alzheimer's disease from EEG
signals, where are we standing?, Current Alzheimer’s Research, 7, 487-505.
279. Hughes JR, Shanmugham S, Wetzel LC, Bellur S, Hughes CA.(1989). The relationship
between EEG changes and cognitive functions in dementia, a study in a VA population.
Clin Electroencephalogr, 20,77–85.
280. Kowalski JW, Gawel M, Pfeffer A, Barcikowska M. (2001). The diagnostic value of EEG
in Alzheimer disease, correlation with the severity of mental impairment. J Clin
Neurophysiol, 18, 570–5.
281. Jeong J., Gore J., Peterson B., (2001). Mutual information analysis of the EEG in patients
with Alzheimer’s disease. Clinical Neurophysiology 112, 827–835.
282. Baker M., Akrofi K.(2008). EEG Patterns in Mild Cognitive Impairment (MCI) Patients.
Open Neuroimag J. 2, 52–55.
283. Besthorn C., Zerfass R. (1997). Discrimination of Alzheimer’s disease and normal aging by
EEG data. Electroencephalography and Clinical Neurophysiology, 103,241–248.
284. 311 Czigler B., Csik´os D., Hidasi Z., Anna Ga´al Z., Csibri E., Kiss E., Salacz P., Moln´ar
M. (2008). Quantitative EEG in early Alzheimer’s disease patients—power spectrum
and complexity features. Int. J. Psychophysiol. 68 (1),75–80.
112
285. Latchoumane C.F.V., Vialatte F.B. (2008) Multiway analysis of Alzheimers disease,
classification based on space-frequency characteristics of EEG time series, World
Congress on Engineering, London, United Kingdom, 2–4 July.
286. van Deursen J.A., Vuurman E.F., Verhey F.R., van Kranen-Mastenbroek V.H., and Riedel
W.J. (2009). Increased EEG gamma band activity in Alzheimer’s disease and mild
cognitive impairment. J Neural Transm. 115(9),1301–11.
287. Coben LA, Danziger W, Storandt M.(1985). A longitudinal EEG study of mild senile
dementia of Alzheimer type, changes at 1 year and at 2.5 years. Electroencephalogr Clin
Neurophysiol, 61, 101–12.
288. Hier DB, Mangone CA, Ganellen R, Warach JD, Van Egeren R, Perlik SJ, Gorelick PB
(1991). Quantitative measurement of delta activity in Alzheimer’s disease. Clin
Electroencephalogr, 22, 178–82.
289. Penttila¨ M, Partanen JV, Soininen H, Riekkinen PJ. (1985). Quantitative analysis of
occipital EEG in different stages of Alzheimer’s disease. Electroencephalogr Clin
Neurophysiol, 60,1–6.
290. Coben LA, Danziger WL, Berg L. (1983). Frequency analysis of the resting awake EEG in
mild senile dementia of Alzheimer type. Electroencephalogr Clin Neurophysiol, 55,
372–80.
291. Prichep LS, John ER, Ferris SH. (2006). Prediction of longitudinal cognitive decline in
normal elderly with subjective complaints using electrophysiological imaging.
Neurobiol Aging. 27 (3), 471–481.
292. Prichep LS, John ER, Ferris SH, Reisberg B, Almas M, Alper K, Cancro R. (1994)
Quantitative EEG correlates of cognitive deterioration in the elderly. Neurobiol Aging.
15 (1), 85–90.
293. Osipova D, Ahveninen J, Kaakkola S, Jaaskelainen IP, Huttunen J, Pekkonen E. (2003)
Effects of scopolamine on MEG spectral power and coherence in elderly subjects. Clin
Neurophysiol. 114 (10), 1902–1907.
113
294. Adler G, Brassen S. (2001) Short-term rivastigmine treatment reduces EEG slow-wave
power in Alzheimer patients. Neuropsychobiology., 43 (4), 273–276.
295. Brassen S, Adler G. (2003) Short-term effects of acetylcholinesterase inhibitor treatment on
EEG and memory performance in Alzheimer patients, an open, controlled trial.
Pharmacopsychiatry. 36 (6), 304–308.
296. Kogan EA, Korczyn AD, Virchovsky RG, Klimovizky S, Treves TA, Neufeld MY. (2001)
EEG changes during long-term treatment with donepezil in Alzheimer’s disease
patients. J Neural Transm. 108 (10), 1167–1173.
297. Grunwald M, Hensel A, Wolf H, Weiss T, Gertz HJ. (2007) Does the hippocampal atrophy
correlate with the cortical theta power in elderly subjects with a range of cognitive
impairment? J Clin Neurophysiol. 24 (1), 22–26.
298. Jelic V, Blomberg M, Dierks T. (1998) EEG slowing and cerebrospinal fluid tau levels in
patients with cognitive decline. Neuroreport. 9 (1), 157–160.
299. Vogel, E.K.& Luck, S.J. (2000). The visual N1 component as an index of a discrimination
process. Psychophysiology, 37, 190-203.
300. Clark, V., Fan, S, & Hillyard, S. (1996). Identification of early visual evoked potential
generators by retinotopic and topographic analyses. Human Brain Mapping, 2, 170-187.
301. Strik, W., Fallgatter, A., Brandeis, D., & Pascual-Marqui, R. (1998). Three-dimensional
tomography of event-related potentials during response inhibition, Evidence for phasic
frontal lobe activation. Electroencephalography & Clinical Neurophysiology, 108, 406-
413
302. Rossion, B., Campanella, S., Gomez, C., Delinte, A,. Debatissem, D., Liard, L., Dubois, S.,
Bruyerm R., Grommelinck, M., & Guerit, J.-M. (1999). Task modulation of brain
activity related to familiar and unfamiliar face processing, an ERP study. Clinical
Neurophysiology, 110, 449-462. P-report
303. Hugdahl, K. (1995). Psychophysiology, The mind-body perspective. Harvard University
Press, Cambridge.
114
304. McCallum, W.C. & Curry, S.H. (1980). The form and distribution of auditory evoked
potentials and CNVs when stimuli and responses are lateralized. Progress in Brain
Research, 54, 767-775.
305. Cecchi, M., Moore, D.K., Sadowsky, C.H. (2015). A clinical trial to validate event-related
potential markers of Alzheimer’s disease in outpatient settings, Alzheimer’s &
Dementia, Diagnosis, Assessment & Disease Monitoring, 1, 387-394.
306. Waldo, M., Gerhardt, G., Baker, N., Drebing, C., Adler, L., & Freedman, R. (1992).
Auditory sensory gating and catecholamine metabolism in schizophrenic and normal
subjects. Psychiatry Research, 44, 21-32.
307. Thoma, R.J., Hanlon, F.M., Moses, S.N., Edgar, J.C., Huang, M., Weisend, M.P., Irwin, J.,
Sherwood, A., Paulson, K., Bustillo, J., Adler, L.E., Miller, G.A., & Canive, J.M.
(2003). Lateralization of auditory sensory gating and neuropsychological dysfunction in
schizophrenia. American Journal of Psychiatry, 160, 1595-1605.
308. Huotilainen, M., Winkler, I., Alho, K., Escera, C., Virtanen, J., Ilmoniemi, R.J.,
Jaaskelainen, I.P., Pekkonen, E., & Naatanen, R. (1998). Combined mapping of human
auditory EEG and MEG responses. Electroencephalography & Clinical
Neurophysiology, 108, 370- 379
309. Potts, G., Dien, J., Hartry-Speiser, A., McDougal, L., & Tucker, D. (1998). Dens sensor
array topography of th event-related potential to task-relevant auditory stimuli.
Electroencephalography and clinical Neurophysiology, 106, 444-456.
310. Luck, S. (1995). Multiple mechanisms of visual-spatial attention, recent evidence from
human electrophysiology. Behavioural Brain Research, 71, 113-123.
311. Mangun, G.R., Hillyard, S.A., & Luck, S.J. (1993). Electrocortical substrates of visual
selective attention. In D. Meyer & S. Kornblum (Eds). Attention and performance 14,
Synergies in experimental psychology, artificial intelligence, and cognitive neuroscience
( 219- 243). MIT Press.
115
312. stimulus Naatanen, R., & Picton, T. W. (1987). The N1 wave of the human electric and
magnetic response to sound, A review and analysis of the component structure.
Psychophysiology, 24, 375- 425.
313. Hillyard, S.A., Hink, R.F., Schwent, V.L., & Picton, T.W. (1973). Electrical signs of
selective attention in the human brain. Science, 182, 177-180.
314. Knight, R.T., Hillyard, S.A., Woods, D.L., & Neville, H.J. (1981). The effects of frontal
cortex lesions on event-related potentials during auditory selective attention.
Electroencephalography and Clinical Neurophysiology, 52, 571-582
315. Ritter, W., Simson, R., & Vaughan, H. (1988). Effects of the amount of stimulus
information processed on negative event-related potentials. Electroencephalography and
clinical Neurophysiology, 69, 244-258.
316. Mangun, G. (1995). Neural mechanisms of visual selective attention. Psychophysiology, 32,
4-18.
317. Vaughan, H.G. Jr. & Ritter, W. (1970). The sources of auditory evoked responses recorded
from the human scalp. Electroencephalography and Clinical Neurophysiology, 28, 360-
367.
318. Papanicolaou, A., Baumann, S., Rogers, R., Saydjari, C., Amparo, E., & Eisenberg, H.
(1990). Localization of auditory response sources using magnetoencephalography and
magnetic resonance imaging. Archives of Neurology, 47, 33-37.
319. Scherg, M., Vajsar, J., & Picton, T. (1989). A source analysis of the late human auditory
evoked potentials. Journal of Cognitive Neuroscience, 1., 336-355
320. Knight, R.T., Scabini, D., Woods, D.L., & Clayworth, C.C. (1988). The effects of lesions of
superior temporal gyrus and inferior parietal lobe on temporal and vertex components of
the human AEP. Electroencephalography & Clinical Neurophysiology, 70, 499-508.
321. Mangun, G.R. & Hillyard, S.A. (1990). Allocation of visual attention to spatial location,
Eventrelated brain potentials and detection performance. Perception and
Psychophysiology, 47, 532-550.
116
322. Coull, J. (1998). Neural correlates of attention and arousal, Insights from electrophysiology,
functional neuroimaging and psychopharmacology. Progress in Neurology, 55, 343-361.
323. Hopf, J.-M., Vogel, E., Woodman, G., Heinze, H.-J., & Luck, S. (2002). Localizing visual
discrimination processes in time and space. Journal of Neurophysiology, 88, 2088-2095
324. Bokura, H., Yamaguchi, S., & Kobayashi, S. (2001). Electrophysiological correlates for
response inhibition in a Go/NoGo task. Clinical Neurophysiology, 112, 2224-2232.
325. Dunn, B. R., Dunn, D. A., Languis, M. & Andrews, D. (1998). The relation of ERP
components to complex memory processing, Brain and Cognition, 36, 355-376. Hyde,
1997
326. Ponton, C.W., Don, M, Eggermont, J.J., Waring, M.D., & Masuda, A. (1996). Maturation of
human cortical auditory function, Differences between normal-hearing children and
children with cochlear implants. Ear and Hearing, 17, 430-437.
327. Hegerl. U. (1993) Intensity dependence of auditory evoked potentials as an indicator of
central serotonergic neurotransmission, A new hypothesis. Biological Psychiatry, 33,
173-187
328. Zouridakis, G., Simos, P.G., & Papanicolaou, A. (1998). Multiple bilaterally asymmetric
cortical sources for the auditory N1m component. Brain Topography , 10, 183-189
329. Hegerl, U., Gallinat J., & Mrowinski, D. (1994). Intensity dependence of auditory evoked
dipole source activity. International Journal of Psychophysiology, 17, 1-13.
330. Scherg, M., & Berg, P. (1991). Use of prior knowledge in brain electromagnetic source
analysis. Brain Topography, 4, 143-150.
331Heslenfeld, D., Kenemans, J.L., Kok, A.,& Molenaar, P. C. M. (1997). Feature processing
and attention in the human visual system, An overview. Biological Psychology, 45, 183-
215.
117
332. van der Stelt, O., Kok, A., Smulders, F.T.Y., Snel, J., & Gunning, B. (1998). Cerebral
event-related potentials associated with selective attention to color, Developmental
changes from childhood to adulthood. Psychophysiology, 35, 227-239.
333. Talsma, D. & Kok, A. (2001). Nonspatial intermodal selective attention is mediated by
sensory brain areas, Evidence from event related potentials. Psychophysiology, 38, 736-
751.
334. Johnson, R. (1989). Developmental evidence for modality dependent P300 generators, A
normative study, Psychophysiology, 26, 651-667.
335. Hackley, S.A., Woldorff, M., & Hillyard, S.A. (1990). Cross-modal selective attention
effects on retinal, myogenic, brainstem, and cerebral evoked potentials.
Psychophysiology, 27, 195- 208.
336. Luck, S.J., & Hillyard, S.A. (1994). Spatial filtering during visual search, evidence from
human electrophysiology. Journal of Experimental Psychology, Human Perception &
Performance, 20, 1000-1014.
337. Novak, G., Ritter, W., & Vaughan Jr., H. (1992). Mismatch detection and the latency of
temporal judgments. Psychophysiology, 29, 398-411.
339. Garcia-Larrea, L., Lukaszewicz, A., & Mauguiere, F. (1992). Revisiting the oddball
paradigm. Non-target vs. neutral stimuli and the evaluation of ERP attentional effects.
Neuropsychologia, 30, 723-741.
340. Sinha V.k., Sur, S. (2009) Event-related potential, An Overview, Industrial Psychiatry
Journal, 18, 70-73.
341. Ritter, W., Simson, R., & Vaughan, H. (1983). Event-related potential correlates of two
stages of information processing in physical and semantic discrimination tasks.
Psychophysiology, 20, 168-179.
342. Satterfield, J. H., Schell, A. M., Nicholas, T. W., Satterfield, B. T., & Freese, T.W. (1990).
Ontogeny of selective attention effects on event-related potentials in attention-deficit
hyperactivity disorder and normal boys. Biological Psychiatry, 28, 879-903.
118
343. Donchin, E. (1978). Use of scalp distribution as a dependent variable in event-related
potential studies, Excerpts of preconference correspondence. In D. Otto (Ed.),
Multidisciplinary perspectives in event-related brain potentials research (501-510),
Washington, DC, US Government Printing Office.
344. Simson, R., Vaughan Jr., H. G., & Ritter, W. (1977). The scalp topography of potentials
inauditory and visual discrimination tasks. Electroencephalography & Clinical
Neurophysiology, 42, 528-604.
345 Bruneau, N. & Gomot, M. (1998). Auditory evoked potentials (N1 wave) as indices of
cortical development. In B. Garreau (Ed.) Neuroimaging in child neuropsychiatric
disorders ( 113-124). Berlin, Springer.
346. Allison, T., Puce, A., Spencer, D., & McCarthy, G. (1999). Electrophysiological studies of
human face perception, I. Potentials generated in occipitotemporal cortex by face and
non-face stimuli. Cerebral Cortex, 9, 415-43.
347. Simson, R., Vaughan Jr., H. G., & Ritter, W. (1977). The scalp topography of potentials in
auditory and visual discrimination tasks. Electroencephalography & Clinical
Neurophysiology, 42, 528-604.
348. Roth, W.T., Kopell, B.S., Tinklenberg, J.R., Huntsberger, G.E., & Kraemer, H.C. (1975).
Reliability of the contingent negative variation and the auditory evoked potential.
Electroencephalography and Clinical Neurophysiology, 65, 45-50.
349. Sandman, C., & Patterson, J. (2000). The auditory event-related potential is a stable and
reliable measure in elderly subjects over a 3 year period. Clinical Neurophysiology, 111,
1427-1437.
350. Shelley, A.M., Ward, P.B., Michie, P. T., Andrews, S., Michell, P.G., Catts, S.V., &
McConaghy, N. (1991). The effect of repeated testing on ERP components during
auditory selective attention. Psychophysiology, 28, 496-510.
119
351. Michalewski, H. H., Prasher, D.K., & Starr, A. (1986). Latency variability and temporal
interrelationship of the auditory event-related potentials (N1, P2, N2, and P3) in normal
subjects. Electroencephalography and Clinical Neurophysiology, 65, 59-71.
352. Pekkonen, E., Rinne, T., & Naatanen, R. (1995). Variability and replicability of the
mismatch negativity. Electroencephalography and Clinical Neurophysiology, 96, 546-
554.
353. Sutton S, Braren M, Zubin J, John E. (1965) Evoked potential correlates of stimulus
uncertainty. Science, 150,1187–8.
354. Polich, J. (2007) Updating P300, An integrative theory of P3a and P3b. Clinical
Neurophysiology, 118, 2128-2148.
355. Bashore T, van der Molen M. (1991) Discovery of P300, a tribute. Biol Psychol, 32, 155–71
356. Donchin, E., Ritter, W., & McCallum, C. (1978). Cognitive psychophysiology, The
endogenous components of the ERP. Brain event-related potentials in man. New York,
NY, Academic Press, 379-411.
357. Callaway P, Tueting P, Koslow S. (1978) Brain-event related potentials in man. New York,
NY, Academic Press, 349–411.
358. Pritchard WS. (1981) Psychophysiology of P300. Psychol Bull, 89,506–40.
359. Donchin E. (1981) Surprise!....Surprise? Psychophysiology, 18, 493–513.
360. Heslenfeld D. (2003) Visual mismatch negativity. In, Polich J, editor. Detection of change,
event-related potential and fMRI findings. Boston, MA, Kluwer, 41–59.
361. Kujala A, Na¨a¨ta¨nen R. (2003) Auditory environment and change detection as indexed by
the mismarch gegativity (MMN). Boston, MA, Kluwer, p. 1–22.
362. Polich J. (1989) Habituation of P300 from auditory stimuli. Psychobiology, 17,19–28.
363. Duncan-Johnson CC, Donchin E. (1977) Effects of a priori and sequential probability of
stimuli on event-related potential. Psychophysiology,14,95.
120
364. Johnson R, Donchin E. (1982) Sequential expectancies and decision-making in a changing
environment, an electrophysiological approach. Psychophysiology,19, 183–200.
365. Squires KC, Wickens C, Squires NK, Donchin E. (1976) Effect of stimulus sequence on
waveform of cortical event-related potential. Science, 193, 1142–6.
366. Snyder E, Hillyard S. (1976) Long-latency evoked potentials to irrelevant, deviant stimuli.
Behav Biol, 16, 319–31.
367. Squires N, Squires K, Hillyard S. (1975) Two varieties of long-latency positive waves
evoked by unpredictable auditory stimuli in man. Electroencephalogr Clin
Neurophysiol, 38,387–401
368. Courchesne E, Hillyard SA, Galambos R. (1975) Stimulus novelty, task relevance, and
visual evoked-potential in man. Electroencephalogr Clin Neurophysiol, 39, 131–43.
369. Knight RT. (1984) Decreased response to novel stimuli after prefrontal lesions in man.
Electroencephalogr Clin Neurophysiol, 59, 9–20.
370. Grunwald T, Lehnertz K, Heinze HJ, Helmstaedter C, Elger CE.(1998) Verbal novelty
detection within the hippocampus proper. Proc Natl Acad Sci USA, 95, 3193–7.
371. Knight RT. (1996) Contribution of human hippocampal region to novelty detection. Nature
, 383, 256–9.
372. Halgren E, Squires N, Wilson C, Rohbraugh J, Bab T, Crandall P. (1980) Endogenous
potentials in the human hippocampal formation and amygdala by infrequent events.
Science, 210, 803–5.
373. McCarthy G, Wood C, Williamson P, Spencer D. (1989) Task-dependent field potentials in
human hippocampal formation. J Neurosci,9,4253–66.
374. Woodman, G. F., & Luck, S. J. (2003). Serial deployment of attention during visual
search. Journal of Experimental Psychology, Human Perception and Performance, 29,
121-138.
121
375. Vespa, Paul M., Nenov, Val, Nuwer, Marc R. (1999). Continuous EEG Monitoring in the
Intensive Care Unit, Early Findings and Clinical Efficacy. Journal of Clinical
Neurophysiology. 16 (1), 1–13.
376. Jain, A., Kim, I. (2011) Low Cost Electroencephalographic Acquisition Amplifier to serve
as Teaching and Research Tool, Engineering in Medicine and Biology Society, 1888-
1891.
377. Mennes, M., Wouters, H. (2010) Validation of ICA as a tool to remove eye movement
artifacts from EEG/ERP, Psychophysiology, 47, 1142-50.
378. Gwin, J.T., Gramann, K. (2010) Removal of movement artifact from high-density EEG
recorded during walking and running, Journal of neurophysiology, 103, 3526-34.
379. Holt, L. E. (1995) P300 topography in Alzheimer's disease. Psychophysiology 32, 257-265.
380. verma , E. R. & Blum, K. (2003). P300 (latency) event-related potential, an accurate
predictor of memory impairment. Clinical EEG, 34, 124-139.
381. Gironell, A., Sanchez, C.G. (2005) Usefulness of P300 in Subjective Memory Complaints,
Journal of Clinical Neurophysiology, 22, 279-284.
382. Papaliagkas, V., Kimiskidis, V., Tsolaki, M. & Anogianakis, G. (2008) Usefulness of event-
related potentials in the assessment of mild cognitive impairment. BMC neuroscience 9,
107.
383. Newsome, R. N., Pun, C.(2013). Neural correlates of cognitive decline in older adults at-
risk for developing MCI, evidence from the CDA and P300. Cognitive neuroscience 4,
152-162.
384. Yener, G. G., Basar, E. (2013). Biomarkers in Alzheimer's disease with a special emphasis
on event-related oscillatory responses. Supplements to Clinical neurophysiology 62,
237-273.
122
385. Jackson, C. E. & Snyder, P. J. (2008). Electroencephalography and event-related potentials
as biomarkers of mild cognitive impairment and mild Alzheimer's disease. Alzheimer's
& dementia , the journal of the Alzheimer's Association 4, S137-143.
386. Lizio, R. (2011). Electroencephalographic rhythms in Alzheimer's disease. International
journal of Alzheimer's disease 2011, 927573.
387. Pedroso, R. V. (2012). P300 latency and amplitude in Alzheimer's disease, a systematic
review. Brazilian journal of otorhinolaryngology 78, 126-132.
389. Patterson, J.V., Michalewski, H.J. (1988) Latency variability of the components of auditory
event-related potentials to infrequent stimuli in aging, Alzheimer-type dementia, and
depression, EEG, 71, 450-460.
390. Kraiuhin C, Gordon E, Coyle S, et al. (1990) Normal latency of the P300 event-related
potential in mild-to-moderate Alzheimer’s disease and depression. Biol Psychiatry
28,372– 86.
391. Johnson, R. (1989). Developmental evidence for modality dependent P300 generators, A
normative study, Psychophysiology, 26, 651-667.
392. Johnson, R. (1993). On the neural generators of the P300 component of the event-related
potential. Psychophysiology, 30, 90-97.
393. Simson, R., Vaughan Jr., H. G., & Ritter, W. (1977). The scalp topography of potentials in
auditory and visual discrimination tasks. Electroencephalography & Clinical
Neurophysiology, 42, 528-604.
394. Pokryszko-Dragan A, Slotwinski K, Podemski R. (2003) Modality-specificchanges in P300
parameters in patients with dementia of the Alzheimer type. Med Sci Monit 9, CR130–4.
395. Papka, M., Ivry, R.B., Woodruff-Pak, D.S. (1997) Eyeblink classical conditioning and
awareness revisited. Psychological Science, 8, 404–408.
396. Woodruff-Pak, D.S., Parka, M. (1996) Eyeblink classical conditioning in Alzheimer’s
Disease and Cerebrovascular Dementia, Neurobiology of Aging, 17, 505-512.
123
397. Solomon, P.R., Levine, E. (1991) Disruption of Classical Conditioning in Patients With
Alzheimer's Disease, Neurobiology of Aging, 12, 283-287.
398. Kaneko T, Thompson RF. (1997). Disruption of trace conditioning of the nictitating
membrane response in rabbits by central cholinergic blockade. Psychopharmacology,
131,161–166.
399. Munera A, Gruart A, Munoz MD, Delgado GJM. (2000). Scopolamine impairs information
processing in the hippocampus and performance of a learned eyeblink response in alert
cats. Neurosci Lett, 292, 33–36.
400. Tanninen, S.E., Yu. X.T., Giritharan, T., Tran, L., Bakir, R., Volle, J., Morrissey, M.D.,
Takehara-Nishiuchi, K. (2015), Cholinergic, but not NMDA, Receptors in the Lateral
Entorhinal Cortex Mediate Acquisition in Trace Eyeblink Conditioning, Hippocampus,
25, 1456-1464.
401. Berger, T. W., Berry, S. D., Thompson, R. F. Role of hippocampus in classical
conditioning of aversive and appetitive behaviors. In, Isaacson, R. L. Pribram, K. L.,
eds. The hippocampus. New York, Plenum Press, 1986.
402. Moore, J. W., Solomon, P. R. Forebrain-brainstem interaction, conditioning and the
hippocampus. In, Butters, N., Squire, L. R., eds. The neuropsychology of memory. New
York, Guilford Press, 1984.
403. Tanninen, S. E., Morrissey, M. D., Klein, R. L., Takehara-Nishiuchi, K. Entorhinal tau
pathology affects local neurons and cortical theta oscillations during memory
acquisition. Alzheimer's & dementia , the journal of the Alzheimer's Association 4, P785
(2014).
404. Buchanan, S.L. and Powell, D.A. (1980) Divergencies in Pavlovian conditioned heart rate
and eyeblink responses produced by hippocampectomy in the rabbit, Behav. Neural
Biol., 30, 20–38.
124
405. Weikart, C.L., Berger, T. (1986) Hippocampal lesions disrupt classical conditioning of
cross-modality reversal learning of the rabbit nictitating membrane response, Behavioral
Brain Research, 22, 85-89.
406. Chachich, M., Powell, D.A. (1998) Both medial prefrontal and amygdala central nucleus
lesions abolish heart rate classical conditioning, but only prefrontal lesions impair
reversal of eyeblink differential conditioning, Neuroscience Letters, 257, 151-154.
407. Shohamy, D., Myers, C.E. (2008) Distinct Hippocampal and Basal Ganglia Contributions to
Probabilistic Learning and Reversal, Journal of Cognitive Neuroscience, 21, 1820–1832.
408. Hunnell, H.K., Boyes, W.K. (1990) Rat and human visual-evoked potentials recorded under
comparable conditions, a preliminary analysis to address the issue of predicting human
neurotoxic effects from rat data. Neurotoxicology and Teratology, 12(4), 391-8.
409. Sambeth, A., Maes, J.H.R. (2006) A comparison of event-related potentials of humans and
rats elicited by a serial feature-positive discrimination task, Learning and Motivation,
37, 269-288.
410. Sambeth, A., Maes, J.H.R. (2003) Auditory event-related potentials in humans and rats,
Effects of task manipulation, Psychophysiology, 40, 60-68
411. Klein, R.L, Dayton, R.D., Tatom, J.B. (2008) Tau expression levels from various adeno-
associated virusvector serotypes produce graded neurodegenerative disease state,
European Journal of Neuroscience, 27, 1615-1625.
412. Sipos, M.L., Burchnell, V., Galbicka, G. (2001). Effects of selected anticholinergics on
acoustic startle response in rats. Journal of Applied Toxicology, 21, 95–101
413. Morrissey MD, Maal-Bared G, Brady S, Takehara-Nishiuchi K (2012). Functional
dissociation within the entorhinal cortex for memory retrieval of an association between
temporally discontinuous stimuli. J Neurosci, 32(16), 5356-5361
414. Chih-Chung Chang and Chih-Jen Lin, (2011). LIBSVM , a library for support vector
machines. ACM Transactions on Intelligent Systems and Technology, 2,27,1--27,27,
Software available at http,//www.csie.ntu.edu.tw/~cjlin/libsvm
125
415. Golub T., Slonim D., Tamayo P., Huard C., Gassenbeek M., Mesirov J., Coller H. (1999)
Modecular Classification of cancer, Class discovery and class prediction by gene
expression monitoring. Science, 286, 531-537.
416. Li, X., Yan, Y. (2013) Identifying Patients with Poststroke Mild Cognitive Impairment by
Pattern Recognition of Working Memory Load-Related ERP, Computational and
Mathematical Methods in Medicine, 1-10.
417. Furey T., Cristianini N., Duffy N., Bednarski D.(2000) Support Vector Machine
Classification and Validation of Cancer Tissue Samples using Microarray Expression
Data, Bioinformatics, 16, 906-914.
418. Sivarajah, Y., Holden, E., Togneri, R. (2014) Quantifying target spotting performances with
complex geoscientific imagery using ERP P300 responses, Human-Computer Studies,
72, 275-283.
419. Ventouras, E.M., Asvestas, P. (2011). Classification of Error-Related Negativity (ERN) and
Positivity (Pe) potentials using kNN and Support Vector Machines, Computers in
Biology and Medicine, 98-108
420. Oloffson, J.K., Ericsson, E., Nordin, S. (2008). Comparison of chemosensory, auditory and
visual event-related potential amplitudes, Scandinavian Journal of Psychology, 49(3),
231-7.
421. Cacioppo, J. T., Tassinary, L. G. & Berndtson, G. (2000). Handbook of psychophysiology.
Cambridge, Cambridge University Press.
422. Braak, H., Del, T.K. (2011). The pathological process underlying Alzheimer’s disease in
individuals under thirty, Acta Neuropathology, 121(2), 171-81.
423. Paxinos G, Watson C. (2007). The rat brain in stereotaxic coordinates. 6th Edition.
Amsterdam , Boston, Academic Press/Elsevier.
424. Liu, L., Drouet, V., Wu, J.W. (2012). Trans-Synaptic Spread of Tau Pathology In Vivo,
PLoS One, 7, 1-9.
126
425. Dujardin S., Lecolle, K. (2014). Neuron-to-neuron wild-type Tau protein transfer through a
trans-synaptic mechanism, relevance to sporadic tauopathies, Acta Neuropathologica
Communications, 2, 1-14.
426. Agster, K., Burwell, R. (2011). Cortical efferents of the perirhinal, postrhinal, and
entorhinal cortices of the rat, Hippocampus, 19, 1159-1186.
427. Kopelman and Corn, 1988 Kopelman, M.D., Corn, T.H. (1988). Cholinergic ‘blockade’ as a
model for cholinergic depletion. Brain 111, 1079–1110.
428. Potter, P., Rauschkolb, P.K. (2011). Pre- and post-synaptic cortical cholinergic deficits are
proportional to amyloid plaque presence and density at preclinical stages of Alzheimer’s
disease, Acta Neuropathology, 122, 49-60.
429. Harris, J. (2012). Human P301L-mutant tau expression in mouse entorhinal-hippocampal
network causes tau aggregation and presynaptic pathology but no cognitive deficits. PloS
one 7.
430. Tanninen, S.E., Morrissey, M.D., Klein, R.L., Takehara-Nishiuchi, K. (2014) Entorhinal tau
pathology affects local neurons and cortical theta oscillations during memory acquisition.
Alzheimer's & dementia: the journal of the Alzheimer's Association, 4, 785.
431. Papp, K., Rentz, D. (2015). The Neuropsychology of Preclinical Alzheimer’s Disease:
Differential Sensitivity of Component Processes of Memory Performance on Biomarker
Evidence of Amyloidosis, Neurology, 84.
432. Isoglu, A.U. Kedzior, K. (2007). Event-related potentials during auditory oddball, and
combined auditory oddball-visual paradigms, International Journal of Neuroscience, 117,
487-506.
433. Sambeth, A., Maes, J.H.R. (2006). A comparison of event-related potentials of humans and
rats elicited by a serial feature-positive discrimination task, Learning and Motivation, 37,
269-288.
434. Lai, C.L., Lin, R.T., Liou, L.M. (2009). The role of event-related potentials in cognitive
decline in Alzheimer’s disease, Clinical Neurophysiology, 121, 194-199.
127
435. Medvidovic, S., Titlic, M., Simunic, M.M. (2013). P300 Evoked Potential in Patients with
Mild Cognitive Impairment, Acta Informatica Medica, 21, 89-92.
436. Hoover BR, Reed MN, Su J, Penrod RD, Kotilinek LA (2010) Tau mislocalization to
dendritic spines mediates synaptic dysfunction independently of neurodegeneration.
Neuron, 68(6), 1067–1081
437. Roy, S., Zhang, B., Lee, V.M., and Trojanowski, J.Q. (2005). Axonal transport defects: a
common theme in neurodegenerative diseases. Acta Neuropathol. 109, 5–13.
438. Shemesh, O.A., Erez, H., Ginzburg, I., and Spira, M.E. (2008). Tau-induced traffic jams
reflect organelles accumulation at points of microtubule polar mismatching. Traffic, 9,
458–471.
439. Gendron TF, Petrucelli L (2009) The role of tau in neurodegeneration. Mol Neurodegener,
4:13
440. Pritchard SM, Dolan PJ, Vitkus A, Johnson GV (2011) The toxicity of tau in Alzheimer
disease: turnover, targets and potential therapeutics. J Cell Mol Med (Epub ahead of
print).
441. Hoover BR, Reed MN, Su J, Penrod RD, Kotilinek LA, (2010) Tau mislocalization to
dendritic spines mediates synaptic dysfunction independently of neurodegeneration.
Neuron, 68(6):1067–1081.
442. McKinney RA (2010) Excitatory amino acid involvement in dendritic spine formation,
maintenance and remodeling. J Physiol, 588(Pt 1):107–116.
443. Rall W (1969). Time constants and electrotonic length of membrane cylinders and neurons.
Biophys J 9(12):1483–1508
444. Caspi-Menkes, N., Yamin, G.G, Kelner, V. (2015) Pathological Tau Disrupts Ongoing
Network Activity, Neuron, 85, 959-966.
128
445. Hoover BR, Reed MN, Su J, Penrod RD, Kotilinek LA. (2010). Tau mislocalization to
dendritic spines mediates synaptic dysfunction independently of neurodegeneration.
Neuron 68(6):1067–1081
446. Roy, S., Zhang, B., Lee, V.M., and Trojanowski, J.Q. (2005). Axonal transport defects: a
common theme in neurodegenerative diseases. Acta Neuropathol. 109, 5–13.
447. Shemesh, O.A., Erez, H., Ginzburg, I., and Spira, M.E. (2008). Tau-induced traffic jams
reflect organelles accumulation at points of microtubule polar mismatching. Traffic, 9,
458–471.
448. Stamer, K., Vogel, R., Thies, E., Mandelkow, E., and Mandelkow, E.M. (2002). Tau blocks
traffic of organelles, neurofilaments, and APP vesicles in neurons and enhances
oxidative stress. J. Cell Biol., 156, 1051–1063.
449. Yoshiyama, Y., Higuchi, M., Zhang, B., Huang, S.M., Iwata, N., Saido, T.C., Maeda, J.,
Suhara, T., Trojanowski, J.Q., and Lee, V.M. (2007). Synapse loss and microglial
activation precede tangles in a P301S tauopathy mouse model. Neuron, 53, 337–351.
450. Yoshiyama Y, Higuchi M, Zhang B, Huang SM, Iwata N, Saido TC, Maeda J, Suhara T,
Trojanowski JQ, Lee VM (2007). Synapse loss and microglial activation precede tangles
in a P301S tauopathy mouse model. Neuron 53(3):337–351
451. Spires TL, Orne JD, SantaCruz K, Pitstick R, Carlson GA, Ashe KH, Hyman BT (2006).
Region-specific dissociation of neuronal loss and neurofibrillary pathology in a mouse
model of tauopathy. Am J Pathol 168(5),1598–1607
452. Crimins, J.L., Rocher, A.B., Peters, A., Shultz, P., Lewis, J., and Luebke, J.I. (2011).
Homeostatic responses by surviving cortical pyramidal cells in neurodegenerative
tauopathy. Acta Neuropathol. 122, 551–564
453. Small DH (2008) Network dysfunction in Alzheimer’s disease: does synaptic scaling drive
disease progression? Trends Mol Med, 14(3):103–108
454. Dudar JD. (1975). The effect of septal nuclei stimulation on the release of acetylcholine
from the rabbit hippocampus. Brain Res, 83:123–33.
129
455. Smith CM. (1974). Acetylcholine release from the cholinergic septohippocampal pathway.
Life Sci, 14:2159–66.
456. Lamour Y, Senut MC, Bassant MH, Dutar P. Functional properties of septo-hippocampal
neurons. In: Steriade M, Biesold D, editors. Brain Cholinergic Systems. Oxford, UK:
Oxford University Press, 1990:265–81.
457. Graef S, Schonknecht P, Sabri O, Hegerl U. (2011). Cholinergic receptor subtypes and their
role in cognition, emotion, and vigilance control: an overview of preclinical and clinical
findings. Psychopharmacology (Berl), 215:205–229.
458. Klinkenberg I, Sambeth A, Blokland A. (2010). Acetylcholine and attention. Behav Brain
Res, 221:430–442. .
459. Murray CL, Fibiger HC. (1985). Learning and memory deficits after lesions of the nucleus
basalismagnocellularis: reversal by physostigmine. Neuroscience. 14:1025–1032.
460. Blake MG, Boccia MM, Krawczyk MC, Baratti CM. (2011). Scopolamine prevents
retrograde memory interference between two different learning tasks. Physiol Behav.,
102:332–337.
461. Klinkenberg I, Blokland A. (2011). A comparison of scopolamine and biperiden as a rodent
model for cholinergic cognitive impairment. Psychopharmacology (Berl)., 215:549–566.
462. Rasmusson, D. D. (2000). The role of acetylcholine in cortical synaptic plasticity.
Behavioural Brain Research, 115, 205–218.
463. Bernardo LS, Prince DA. (1982). Cholinergic excitation of mammalian hippocampal
pyramidal cells. Brain Res, 249:315–31.
464. Cole AE, Nicoll RA. ( 1984). Characterization of a slow cholinergic post-synaptic potential
recorded in vitro from rat hippocampal pyramidal cells. J Physiol, 352:173–88.
465. Halliwell JV, Adams PR. (1982). Voltage-clamp analysis of muscarinic excitaiton in
hippocampal neurons. Brain Res, 250:71–92.
130
466. Abe K, Nakata A, Mizutani A, Saito H. (1994) Facilitatory but nonessential role of
muscarinic cholinergic system in the generation of long-term potentiation of population
spikes in the dendritic gyrus in vivo. Neuropharmacology, 33:847–52.
467. Hirotsu I, Hori N, Katsuda N, Ishihara T. (1989) Effect of anticholinergic drug on long-term
potentiation in rat hippocampal slices. Brain Res, 482:194–7.
468. Auerbach JM, Segal M. (1994) A novel cholinergic induction of long-term potentiation in
rat hippocampus. J Neurophysiol, 72:2034–40.
469. Sarter M, Hasselmo ME, Bruno JP, Givens B. (2005) Unraveling the attentional functions
of cortical cholinergic inputs: interactions between signal-driven and cognitive
modulation of signal detection. Brain Res Rev., 48:98–111.
470. Everitt BJ, Robbins TW. (1997) Central cholinergic systems and cognition. Annu Rev
Psychol., 48:649–684.
471. Furey ML, Pietrini P, Haxby JV. (2000) Cholinergic enhancement and increased selectivity
of perceptual processing during working memory. Science, 290:2315–2319.
472. Sillito AM, Kemp JA. (1983) Cholinergic modulation of the functional organization of the
cat visual cortex. Brain Res. 289 (1–2):143–155.
473. Sato H, Hata Y, Masui H, Tsumoto T. (1987) A functional role of cholinergic innervation to
neurons in the cat visual cortex. J Neurophysiol, 58:765–780.
474. Metherate R., Ashe J.H. (1995) Synaptic interactions involving acetylcholine, glutamate,
and GABA in rat auditory cortex, Exp Brain Res, 107, 59-72.
475. Ashe JH, McKenna TM, Weinberger NM (1989) Cholinergic modulation of frequency
receptive fields in auditory cortex, Synapse, 4, 30-43.
476. Metherate R, Ashe JH (1993) Nucleus basalis stimulation facilitates thalamocortical
synaptic transmission in rat auditory cortex. Synapse 14:132-143
131
477. Ashe, J.H., MCkenna, T.M. (1989) Cholinergic modulation of Frequency-specific effects of
anticholinesterases provide evidence for a modulatory action of endogenous ACh.
Synapse 4:44-54
478. Metherate R, Weinberger NM (1990) Cholinergic modulation of responses to single tones
produces tone-specific receptive field alterations in cat auditory cortex. Synapse 6:133-
145
479. Goldman-Rakic PS. (1995) Cellular basis of working memory. Neuron, 14:477-485.
480. Laroche S, Davis S, Jay TM. (2000) Plasticity at hippocampal to prefrontal cortex synapses:
dual roles in working memory and consolidation, Hippocampus, 10, 438-46.
481. Dash PK, Moore AN, Kobori N, Runyan JD. (2007). Molecular activity underlying working
memory. Learn Mem. 14:554--563.
482. Sesack SR, Deutch AY, Roth RH, Bunney BS. (1989). Topographical organization of the
efferent projections of the medial prefrontal cortex in the rat: an anterograde tract-
tracing study with Phaseolus vulgaris leucoagglutinin. J Comp Neurol. 290:213--242.
483. Conde F, Maire-Lepoivre E, Audinat E, Crepel F. (1995). Afferent connections of the
medial frontal cortex of the rat 2: Cortical and subcortical afferents. J Comp Neurol.
352:567--593.
484. Vertes RP. (2004). Differential projections of the infralimbic and prelimbic cortex in the rat.
Synapse. 51:32--58.
485. Gabbott PL, Warner TA, Jays PR, Salway P, Busby SJ. (2005). Prefrontal cortex in the rat:
projections to subcortical autonomic, motor, and limbic centers. J Comp Neurol.
492:145--177.
486. Steriade M, Nunez A, Amzica F. (1993). Intracellular analysis of relations between the slow
( < 1 Hz) neocortical oscillation and other sleep rhythms of the electroencephalogram. J
Neurosci. 13:3266--3283.
132
487. Steriade M, Nunez A, Amzica F. (1993). A novel slow ( <1 Hz) oscillation of neocortical
neurons in vivo: depolarizing and hyperpolarizing components. J Neurosci. 13:3252--
3265.
488. Jay TM, Glowinski J, Thierry AM. (1989). Selectivity of the hippocampal projection to the
prelimbic area of the prefrontal cortex in the rat. Brain Res. 505:337--340.
489. Jay TM, Witter MP. (1991). Distribution of hippocampal CA1 and subicular efferents in the
prefrontal cortex of the rat studied by means of anterograde transport of Phaseolus
vulgaris-leucoagglutinin. J Comp Neurol. 313:574—586
490. Delatour B, Witter MP. (2002). Projections from the parahippocampal region to the
prefrontal cortex in the rat: evidence of multiple pathways. Eur J Neurosci. 15:1400--
1407.
491. Valenti, O., Grace, A.A. (2009) Entorhinal Cortex Inhibits Medial Prefrontal Cortex and
Modulates the Activity States of Electrophysiologically Characterized Pyramidal
Neurons In Vivo, Cerebral Cortex, 19, 658-674.
492. Saper, C.B. (1984) Organization of cerebral cortical afferent systems in the rat. I.
Magnocellular basal nucleus. J. Comp. Neurol., 222, 313–342.
493. Luiten, P.G.M., Gaykema, R.P.A., Traber, J. & Spencer, D.G. (1987) Cortical projection
patterns of magnocellular basal nucleus subdivisions as revealed by anterogradely
transported Phaseolus vulgaris leucoagglutinin. Brain Res., 413, 229–250.
494. Fanselow, E.E., Richardson, K.A. (2008). Selective, state-dependent activation of
somatostatin-expressing inhibitory interneurons in mouse neocortex. J Neurophysiol,
100, 2640–2652.
495. Pi, H.J., Hangya, B. (2013). Cortical interneurons that specialize in disinhibitory control.
Nature, 503, 521–524.
496. DeFelipe, J. (1997) Types of neurons, synaptic connections and chemical characteristics of
cells immunoreactive for calbindin-D28K, parvalbumin and calretinin in the neocortex.
J. Chem. Neuroanat., 14, 1–19.
133
497. Beaulieu, C. & Somogyi, P. (1991) Enrichment of cholinergic synaptic terminals on
GABAergic neurons and coexistence of immunoreactive GABA and choline
acetyltransferase in the same synaptic terminals in the striate cortex of the cat. J. Comp.
Neurol., 304, 666–680.
498. Nguyen, D.P., Lin, S.C. (2014) A frontal cortex event-related potential driven by the basal
forebrain, eLIFE, 1-17.
499. Freund TF, Meskenaite V. (1992). gamma-Aminobutyric acid-containing basal forebrain
neurons innervate inhibitory interneurons in the neocortex. Proceedings of the National
Academy of Sciences of the United States of America 89:738–742.
500. Gritti I, Mainville L, Mancia M, Jones BE. (1997). GABAergic and other noncholinergic
basal forebrain neurons, together with cholinergic neurons, project to the mesocortex
and isocortex in the rat. The Journal of Comparative Neurology 383:163–177.
501. Henny P, Jones BE. (2008). Projections from basal forebrain to prefrontal cortex comprise
cholinergic, GABAergic and glutamatergic inputs to pyramidal cells or interneurons.
European Journal of Neuroscience 27:654–670.
502. Steele, V.R., Fink, J., Maurer, M. (2014) Brain Potentials Measured During a Go/NoGo
Task Predict Completion of Substance Abuse Treatment, Biological Psychiatry, 76, 75-
83.
503. Munoz, A.R., Lopez, M.M. (2014) EEG study on affective valence elicited by novel and
familiar pictures using ERD/ERS and SVM-RFE, Medical and Biological Engineering
& Computing, 52, 149-158.
504. Cecere, C., Corrado, C., Polikar, R. (2014) Diagnostic Utility of EEG Based Biomarkers for
Alzheimer's Disease, Annual Northeast Bioengineering Conference, DOI:
10.1109/NEBEC.2014.6972751.
505. Ramaniharan, A.K., Monaharan, S.C. (2016) Laplace Beltrami eigen value based
classification of normal and Alzheimer MR images using parametric and non-parametric
classifiers, Expert Systems with Application, 59, 208-216.
134
506. Wei, R.Z., Li, C.H. (2016) Prediction of Conversion from Mild Cognitive Impairment to
Alzheimer's Disease Using MRI and Structural Network Features, Frontiers in Aging
Neuroscience, 8, DOI: 10.3389/fnagi.2016.00076.
507. Colloby, S.J., Cromarty, R.A. Peraza, L.R. (2016) Multimodal EEG-MRI in the differential
diagnosis of Alzheimer's disease and dementia with Lewy bodies, Journal of Psychiatric
Research, 78, 48-55.
508. Vaughan, H.G. Jr. & Ritter, W. (1970). The sources of auditory evoked responses recorded
from the human scalp. Electroencephalography and Clinical Neurophysiology, 28, 360-
367.
509. Papanicolaou, A., Baumann, S., Rogers, R., Saydjari, C., Amparo, E., & Eisenberg, H.
(1990). Localization of auditory response sources using magnetoencephalography and
magnetic resonance imaging. Archives of Neurology, 47, 33-37.
510. Scherg, M., Vajsar, J., & Picton, T. (1989). A source analysis of the late human auditory
evoked potentials. Journal of Cognitive Neuroscience, 1., 336-355.
511. Knight, R.T., Scabini, D., Woods, D.L., & Clayworth, C.C. (1988). The effects of lesions of
superior temporal gyrus and inferior parietal lobe on temporal and vertex components of
the human AEP. Electroencephalography & Clinical Neurophysiology, 70, 499-508.
512. Ashiskali, M., Polikar, R., Kounios, J., Green, D., Clark, C.M. (2009) Combining
Multichannel ERP Data for Early Diagnosis of Alzheimer’s Disease, Engineering in
Medicine & Biology Society, 522-525.
513. Chapman, M.R., Nowlis, G.H., McCrary, J.W. (2007) Brain event-related potentials:
Diagnosing early-stage Alzheimer’s disease, Neurobiology of Aging, 28, 194-201.
514. Papaliagkas, V., Kimiskidis, V., Tsolaki, M. (2008) Usefulness of event-related potentials in
the assessment of mild cognitive impairment, BMC Neuroscience, 107 (9)
515. Chapman, M.R., McCrary, J.W. (2011) Brain ERP components predict which individuals
progress to Alzheimer’s disease and which do not, Neurobiology of Aging, 32, 1742-1755.
135
516. Flor, H., Birbaumer, N., Roberts, L.E. (1996) Slow potentials, event-related potentials,
"gamma-band" activity, and motor responses during aversive conditioning in humans,
Experimental Brain Research, 112, 298-312.
517. Nakamura M, Ozawa N, Shinba T, Yamamoto K (1993) CNV-like potentials on the cortical
surface associated with conditioningin head-restrained rats. Electroencephalogr Clin
Neurophysiol 88:155-162
518. Begleiter H, Platz A (1969) Evoked potentials: modifications by classical conditioning.
Science 166, 769-771
519. Moyer Jr., J.R., Deyo, R.A., and Disterhoft, J.F. (1990). Hippocampectomy disrupts trace
eye-blink conditioning in rabbits. Behav. Neurosci. 104, 243–252.
520. Tseng W., Guan, R., Disterhoft, J.F., and Weiss, C. (2004). Trace eyeblink conditioning is
hippocampally dependent in mice. Hippocampus 14, 58–65.
521. Tanninen, S.E., Morrissey, M.D., Takehara, K. (2013) Unilateral Lateral Entorhinal
Inactivation Impairs Memory Expression in Trace Eyeblink Conditioning, PLoS One, 8
(12), 1-8.
522. Altafullah, I., Halgren, E. (1986). Interictal spike-wave complexes in the human medial
temporal lobe: typical topography and comparisons with cognitive potentials.
Electroencephalogr Clin Neurophysiol, 63:503-516 ,
523. Flor, H., Birbaumer, N., Roberts, L.E. (1996). Slow potentials, event-related potentials,
"gamma-band" activity, and motor responses during aversive conditioning in humans,
Experimental Brain Research, 112, 298-312.
524. Polich, J. (2007). Updating P300: An integrative theory of P3a and P3b, Cilinican
Neurophysiology, 118, 2128-2148.
525. O’Mahony D, Coffey J, Murphy J, OHare N, Hamilton D, Rowan M. (1996). Event-related
potential prolongation in Alzheimers disease signifies frontal lobe impairment: evidence
from SPECT imaging. J Gerontol A Biol Sci Med Sci,. 51(3):102-7.
136
526. Ball SS, Marsh JT, Schubarth G, Brown WS, Strandburg R. (1989). Longitudinal P300
latency changes in Alzheimer’s disease. J Gerontol, 44(6), M195-200.
527. Polich, J., Pitzer, A. (1999). P300 in early Alzheimer's disease: oddball task difficulty and
modality effects. Clinical Neurophysiology: from Receptors To Perception, EEG
Supplement. Elsevier, Amsterdam, pp. 281–287.
528. Ford, J.M., Woodward, S.H., Sullivan, E.V., Isaacks, B.G., Tinklenberg, J.R., Yesavage,
J.A., Roth, W.T., (1996). N400 evidence of abnormal responses to speech in Alzheimer's
disease. Electroencephalogr. Clin. Neurophysiol. 99, 235–246.
529. Holt, L., Raine, A., Pa, G., Schneider, L., Henderson, V., Pollock, V.E. (1995). P300
topography in Alzheimer's disease. Psychophysiology 32, 257–265.
530. Kok A. (1989). On the utility of P3 amplitude as a measure of processing capacity.
Psychophysiology, 38:557–77.
531. Kutas M, McCarthy G, Donchin E. (1977). Augmenting mental chronometry: P300 as a
measure of stimulus evaluation time. Science, 197:792–5.
532. Polich J. (1987).Task difficulty, probability, and inter-stimulus interval as determinants of
P300 from auditory stimuli. Electroencephalogr Clin Neurophysiol, 68:311–20.
533. Davis,H. (1964). Enhancement of evoked cortical potentials in humans related to a task
requiring a decision, Science, 145, 182-183.
534. Donchin.E., Cohen.L. (1967). Averaged evoked potentials and intramodally selective
attention, Electroencephalogr. Clin. Neurophysiol., 22, 537-546,
535. Donchin.E., Lindsley.D.B.(1966). Averaged evoked potentials and reaction times to visual
stimuli, Electroencephalogr. Clin. Neurophysiol., 20, 217-223.
536. Eason.R.G., Harter.M.R.(1969). Effects of attention and arousal on visually evoked cortical
potentials and reaction time in man, Physiol. Behav., 4, 283-28.
137
537. Garcia-Aust.E. (1963). Influence of the states of awareness upon sensory-evoked potentials,
Electroencephalogr. Clin. Neurophysiol., Suppl. 24.
538. Keating, L.W., Ruhm, H.B. (1971). Some observations on the effects of Attention on
Stimuli on the Amplitude of the Acoustically Evoked Response, Audiology, 10, 177-
184.
539. Celebi, G. (1981). The role of stimulus modality and response condition on human event
related potentials, Chemical Senses, 6, 1-8.
540. Dustman,R.E., Beck.E.C. (1965). Phase of alpha brain waves, reaction times and visually
evoked potentials, Electroencephalogr. Clin. Neurophysiol., 18, 433-440.
541. Haider.M., Spong.P. and Lindsley.D.B. (1964). Attention, vigilance and cortical evoked
potentials in humans, Science, 145, 180-182.
542. Bostock.H., Jarvis.M.J. (1970), Changes in the form of the cerebral evoked responses
related to the speed of simple reaction time', Electroencephalogr. Clin. Neurophysiol.,
26, 266-272.
543. Chalupa.L.M., Rohrbaugh.J.W., Gould.J.E., Lindsley.D.B. (1976) Cortical and sub-cortical
visual evoked potential correlates of reaction time in monkeys, .Comp. Physiol.
Psychol., 90, 119-126.
544. Eason.R.G. Harter.M.R (1969). Effects of attention and arousal on visually evoked cortical
potentials and reaction time in man, Physiol. Behav., 4, 283-289.
545. Celebi, G. (1981) The role of stimulus modality and response condition on human event
related potentials, Chemical Senses, 6, 1-8.
546. Polich, J. (2007) Updating P300: An integrative theory of P3a and P3b, Cilinican
Neurophysiology, 118, 2128-2148.
547. Cohen, R.M., Rezai-Zadeh, K., Weitz, T.M. (2013) A transgenic Alzheimer rat with
plaques, tau pathology, behavioral impairment, oligomeric aβ, and frank neuronal loss,
The Journal of Neuroscience, 33, 6245-56.
138
548. Sage Boyertown facility (2015, March 26) Humanized ApoE4 Knock-in Rat By
Application. Retrieved from https://www.horizondiscovery.com/hapoe4-knock-in-rat-
tgra8960
