University of Southern California Dissertations and Theses

Averting dementia: renin-angiotensin system and angiogenic cells in cognitive decline

(USC Thesis Other)

Averting dementia: renin-angiotensin system and angiogenic cells in cognitive decline

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Averting Dementia:
Renin-Angiotensin System
and Angiogenic Cells in Cognitive Decline

by Alick Tan

University of Southern California

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
Clinical and Experimental Therapeutics

August 2017

Dissertation Committee:
Dr. Kathleen E. Rodgers, Advisor and Chair
Dr. Daniel A. Nation, Co-advisor
Dr. Daryl Davies

ii
Dedication

Dedicated to the loving memory of my Grandma (1930–2016)
iii
Acknowledgements
I could not have done any of this work without the guidance and support of my mentor,
Dr. Kathleen Rodgers, along with all the help from the wonderful individuals in her lab:
Sachin Jadhav, Lila Kim, Ania Papińska, Roslynn Stone, Maira Soto, Kevin Gaffney,
Maria Banos, Tamar Amzaleg, Theresa Espinoza, Norma Roda, Chris Meeks,
Rosemary Flores, Josh Dorst, and Michael Weinberg. To the members of my qualifying
and dissertation committees, Drs. Stan Louie, Enrique Cadenas, Daryl Davies, and
Daniel Nation, I thank you for all the support and direction over the years. The clinical
studies, which involve so much work—the participant recruitment, neuropsych, and
MRIs—would not be possible without the collaborative effort of the dedicated individuals
in the Nation lab, including: Elissa McIntosh, Shubir Dutt, Belinda Yew, Jean Ho, Anna
Blanken, Jung Jang, and Aimée Gaubert. I also thank the individuals whom have given
me words of wisdom and encouragement, and lent a helping hand, including: Ishan
Patil, Aarti Mishra, Ron Irwin, and Rachel Reyes. Furthermore, I would be ever lost in
the realm of neuroscience and the clinical world if it were not for Drs. Michael Jakowec,
Bradley Williams, Helena Chui, and Freddi Segal-Gidan showing me the way. All these
individuals, and so many more, have taught me so much by example and so many
different ways.

To everyone, thank you.
iv
Table of Contents
Dedication ....................................................................................................................... ii
Acknowledgements ....................................................................................................... iii
Table of Contents .......................................................................................................... iv
List of Figures ............................................................................................................... xv
List of Tables .............................................................................................................. xvii
List of Abbreviations ................................................................................................. xviii
Abstract ......................................................................................................................... xx
Chapter 1: Introduction & Background ........................................................................ 1
Alzheimer's Disease ..................................................................................................... 1
Pathology .................................................................................................................. 1
Epidemiology ............................................................................................................. 3
Hypotheses of Alzheimer's Disease .......................................................................... 3
Clinical Presentation ................................................................................................. 6
Memory Impairments ............................................................................................. 7
Executive Dysfunction ........................................................................................... 8
Clinical Evaluation ................................................................................................. 9
Neuropsychological Testing ............................................................................. 10
Measuring Clinical Disease Progression ......................................................... 10
Neuroimaging ................................................................................................... 11
Structural Neuroimaging ............................................................................... 12
Structural MRI Findings in Cognitive Decline ............................................... 13
v
Functional and Molecular Neuroimaging ...................................................... 14
Biomarkers ....................................................................................................... 15
Current Therapies ................................................................................................... 17
Cholinesterase Inhibitors ..................................................................................... 19
NMDA Antagonists .............................................................................................. 19
Risk Factors for Cognitive Decline & Dementia .......................................................... 20
Age .......................................................................................................................... 20
Genetics .................................................................................................................. 20
Mild Cognitive Impairment ....................................................................................... 21
Vascular Contributions to Cognitive Impairment and Dementia .............................. 21
Hypertension ....................................................................................................... 22
Cerebrovascular Reactivity .................................................................................. 23
Animal Models of AD .................................................................................................. 25
Behavioral Tasks ..................................................................................................... 26
Renin-Angiotensin System ......................................................................................... 28
Angiotensin enzymes act as Aβ-degrading enzymes ............................................. 31
Vascular Homeostasis and RAS Therapy ............................................................... 32
Brain RAS ............................................................................................................... 33
Cardiovascular and Central Pressor Actions of Angiotensins ............................. 34
Neuroanatomical and Cellular Sources of Angiotensins ..................................... 34
Local Production of Angiotensin Elements ...................................................... 35
Angiotensinogen and Renin ............................................................................. 36
vi
Enzyme Distribution ............................................................................................. 38
ACE .................................................................................................................. 38
ACE2 ................................................................................................................ 39
Aminopeptidases A and N ............................................................................... 40
Ligand Distribution ............................................................................................... 42
Receptor Distribution ........................................................................................... 42
AT1 and AT2 Receptors .................................................................................. 42
AT4 Receptor ................................................................................................... 43
Mas Receptor ................................................................................................... 44
Centrally Active RAS Drugs ................................................................................. 46
Brain RAS Summary............................................................................................ 47
RAS Dysregulation in AD ............................................................................................ 49
Clinical Studies of ARBs in AD ................................................................................ 50
Preclinical Studies of ARBs in AD ........................................................................... 51
Ang II in AD ............................................................................................................. 53
Ang-(1–7) in AD ...................................................................................................... 55
Ang-(1–7) in Learning & Memory ........................................................................ 55
Preclinical Studies of Ang-(1–7) in AD ................................................................. 56
Clinical Studies of Ang-(1–7) in AD ..................................................................... 58
RAS and PPAR-γ Activity ....................................................................................... 60
Endothelial Progenitor Cells ....................................................................................... 62
Methods of Isolation, Enumeration, and Proliferation ............................................. 62
vii
EPCs in AD ............................................................................................................. 65
Effects of Modulating the RAS on EPCs ................................................................. 67
Progenitor Cells in AD ................................................................................................ 69
The Hematopoietic System ..................................................................................... 69
Blood Cell Lineages ............................................................................................. 70
Identifying Progenitors by Cell Surface Antigens ................................................ 71
Progenitor Sources and Uses .............................................................................. 71
Role of RAS in Hematopoiesis ............................................................................ 72
Hematopoietic Growth Factors ............................................................................ 73
Non-hematopoietic Cells in the BM ..................................................................... 74
Cognitive Decline in Disorders of Hematopoietic Deficit ..................................... 75
Adult Neurogenesis ................................................................................................. 76
Neurogenic Niches .............................................................................................. 76
Progenitor Cell Homing to the Brain .................................................................... 77
Clinical Studies of Progenitor Cells in AD ............................................................... 78
Gaps in Knowledge ..................................................................................................... 80
Chapter 2: Angiotensin-(1–7) and Candesartan Reduce Aβ .................................... 82
Introduction ................................................................................................................. 82
Methods ...................................................................................................................... 84
Animals.................................................................................................................... 84
Chemicals and Reagents ........................................................................................ 84
Study Design and Timeline ..................................................................................... 84
viii
Behavioral Tasks ..................................................................................................... 85
Spontaneous Alternations on the T-maze ........................................................... 85
Novel Object Recognition .................................................................................... 86
Prepulse Inhibition ............................................................................................... 87
Carotid Blood Flow .................................................................................................. 87
Tissue Collection ..................................................................................................... 88
EPC Flow Cytometry ............................................................................................... 88
Hippocampal Aβ Immunoassay .............................................................................. 89
Statistical Analysis .................................................................................................. 90
Results ........................................................................................................................ 91
Working Memory Tests: T-maze & Novel Object Recognition ................................ 91
Prepulse Inhibition ................................................................................................... 94
Candesartan reduced carotid blood flow, except in NonTg female mice ................ 96
Candesartan increased circulating Flk1+Sca1+ EPCs ........................................... 97
Candesartan decreased hippocampal insoluble Aβ species .................................. 98
Candesartan decreased survival of female 3xTg-AD mice, rescued by Ang-(1–7)
.............................................................................................................................. 100
Discussion ................................................................................................................ 101
Behavioral Tasks ................................................................................................... 101
Vascular Mechanisms ........................................................................................... 104
Carotid Blood and Survival ................................................................................ 104
EPCs .................................................................................................................. 105
ix
Reduction of Hippocampal Aβ .............................................................................. 106
AT1 Blockade vs. Mas Stimulation ........................................................................ 108
Summary ............................................................................................................... 110
Chapter 3: Aβ Clearance Mechanisms ..................................................................... 112
Introduction ............................................................................................................... 112
Methods .................................................................................................................... 116
mRNA Expression ................................................................................................. 116
Soluble APPα/β Immunoassay ............................................................................. 119
Statistical Analyses ............................................................................................... 119
Results ...................................................................................................................... 120
APP Processing .................................................................................................... 120
Soluble APPα and APPβ ...................................................................................... 121
Aβ-degrading Enzymes ......................................................................................... 122
Glial Activation ...................................................................................................... 123
PPAR-γ/LXR Activation ......................................................................................... 124
Renin-Angiotensin System .................................................................................... 125
Discussion ................................................................................................................ 127
Candesartan modulated APP processing genes Bace1 and Psen1 ..................... 127
Candesartan and Ang-(1–7) did not affect amyloidogenic APP processing ......... 129
Candesartan increased the mRNA expression of Aβ-degrading enzymes: IDE,
ECE2, and ACE2 .................................................................................................. 130
Candesartan upregulated microglial activation genes .......................................... 132
x
PPAR-γ Activation ............................................................................................. 134
Candesartan modulates central RAS receptor expression ................................... 134
Summary ............................................................................................................... 139
Chapter 4: EPC Method Development and RAS Receptor Expression ................. 140
Introduction ............................................................................................................... 140
Methods .................................................................................................................... 143
Peripheral Blood Assays ....................................................................................... 143
Blood Smears: WBC Differentials ...................................................................... 143
Flow Cytometry .................................................................................................. 143
Gating Strategy .............................................................................................. 144
Progenitor Cell Enrichment for Fluorescence-Minus-One Bounds ................ 145
Colony-Forming Assays .................................................................................... 146
CFU-Hill Assay ............................................................................................... 146
Circulating Angiogenic Cells (CFU-EPCs) ..................................................... 147
Functional Assays .......................................................................................... 147
Senescence-associated β-galactosidase staining ..................................... 147
AcLDL Uptake ............................................................................................ 148
Boyden chamber chemotaxis assay ........................................................... 148
Matrigel Tube Formation Assay ................................................................. 148
Statistical Analyses ............................................................................................ 149
Participant Characteristics ............................................................................. 149
Factors Influencing Angiogenic Cell Subsets ................................................. 150
xi
Correlations to CFU-Hill .............................................................................. 150
RAS receptor expression on angiogenic cell subsets ................................ 150
Results ...................................................................................................................... 152
EPC Method Development .................................................................................... 152
Progenitor Enrichment & FMO Bounds ............................................................. 153
Colony-Forming Assays .................................................................................... 155
Functional Assays ............................................................................................. 155
β-Gal .............................................................................................................. 155
AcLDL uptake requires 3–4 hours incubation ................................................ 156
CFU-Hill colonies readily take up AcLDL ....................................................... 158
Correlations to CFU-Hill ........................................................................................ 159
CD34+CD133+ cells correlated with CFU-Hill colonies ..................................... 159
CFU-Hill colonies did not correlate with Framingham risk scores ..................... 160
CFU-Hill colonies did not correlate with hematological parameters .................. 161
CD34+CD133+CD309+ cells express CD34
bright
population ................................. 162
RAS receptor expression on angiogenic cell subsets ........................................... 163
Discussion ................................................................................................................ 165
Assay Development Considerations ..................................................................... 165
Functional Assays ................................................................................................. 166
CD34+CD133+ cells correlated to CFU-Hill colonies ............................................ 168
Correlation of CFU-Hill to other parameters .......................................................... 169
EPC Maturation Stages and RAS Receptor Expression ....................................... 169
xii
Summary ............................................................................................................... 173
Chapter 5: Clinical Study of Angiogenic Cells in MCI ............................................. 174
Introduction ............................................................................................................... 174
Methods .................................................................................................................... 175
Participant Recruitment ......................................................................................... 175
Neuropsychological Assessment .......................................................................... 175
Mild Cognitive Impairment ................................................................................. 177
Participant Characteristics .................................................................................... 177
MRI Volumetry ....................................................................................................... 177
Voxel-based Morphometry and Regression ...................................................... 178
Peripheral Blood Assays ....................................................................................... 178
Plasma Biomarkers............................................................................................ 179
APOE Genotyping ............................................................................................. 180
Flow Cytometry .................................................................................................. 180
Gating Strategy .............................................................................................. 181
CFU-Hill Assay .................................................................................................. 181
Functional Assays .......................................................................................... 182
Senescence-associated β-galactosidase staining ..................................... 182
AcLDL Uptake ............................................................................................ 182
Statistical Analyses and Outcomes ....................................................................... 183
Primary Outcomes: Blood-based Assays in Cognition and Brain Atrophy ........ 183
xiii
Definition of Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers
....................................................................................................................... 183
Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers in Normal
Cognition and MCI ......................................................................................... 184
Correlation of Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers
to Brain Volumes ............................................................................................ 184
Hippocampal Volumetry in MCI .................................................................. 185
Vascular Senescence and Endothelial Dysfunction ....................................... 185
Results ...................................................................................................................... 186
Participant Characteristics .................................................................................... 186
Detection of pro-angiogenic progenitor cell subsets and plasma biomarkers ....... 188
Primary Outcomes ................................................................................................. 188
Initial Sensitivity Analyses ................................................................................. 188
Reduction in CD34 and CD133 Subsets in MCI After Adjusting for Age, Sex, and
Education ........................................................................................................... 191
Patients with higher CD34+ cells performed better on logical memory and visual
reproduction tests .............................................................................................. 195
CD34 cells correlated to voxel-based volumes in brain regions responsible for
visual and verbal memory .................................................................................. 196
Hippocampal volumes did not correlate to angiogenic cell subsets or plasma
biomarkers ......................................................................................................... 197
Hippocampal volume was not significantly reduced in MCI group ................. 197
xiv
Vascular Senescence ........................................................................................ 197
SA-β-Gal in MCI ............................................................................................. 197
Correlation between SA-β-gal and ICAM-1 ................................................... 198
Discussion ................................................................................................................ 199
CD34 deficits in MCI correlated with deficits in verbal and visual memory tests and
voxel-wise regression of MRI brain volumes involved with verbal and visual
memory ................................................................................................................. 199
Identity, significance, and association of progenitor cells in AD ............................ 200
Role of these pro-angiogenic progenitor cells in cognitive decline ....................... 201
Vascular risk and participant population ............................................................... 202
Despite the small sample size, this significant finding correlated with both
neuropsychological & MRI measures .................................................................... 202
Vascular senescence ............................................................................................ 203
Summary ............................................................................................................... 206
Chapter 6: Conclusions and Future Directions ....................................................... 207
References .................................................................................................................. 211
xv
List of Figures
Figure 1 Jack model of AD pathological cascade. ........................................................... 2
Figure 2 Simplified angiotensin pathway. ....................................................................... 30
Figure 3 Pathways of Ang-(1–7) generation. .................................................................. 31
Figure 4 Brain RAS and BBB Permeability. ................................................................... 36
Figure 5 Alternative Ang II Processing by Aminopeptidases. ......................................... 40
Figure 6 Overview of Brain RAS Pathways. ................................................................... 45
Figure 7 Sites of Hematopoiesis in Human Development. ............................................. 70
Figure 8 Animal Study Design and Timeline. ................................................................. 85
Figure 9 EPC Gating Methodology ................................................................................. 89
Figure 10 Spontaneous alternations in T-maze. ............................................................ 92
Figure 11 Novel Exploration Indices. .............................................................................. 93
Figure 12 Total Exploration Time. .................................................................................. 94
Figure 13 Acoustic Startle. ............................................................................................. 95
Figure 14 Prepulse Inhibition. ......................................................................................... 95
Figure 15 Carotid Blood Flow. ........................................................................................ 96
Figure 16 Candesartan increased circulating Flk1+Sca1+ EPCs. ................................. 97
Figure 17 Candesartan decreased hippocampal insoluble Aβ species. ........................ 99
Figure 18 Survival Curve of Female 3xTg-AD Mice. .................................................... 100
Figure 19 Aβ Production and Clearance Pathways. .................................................... 115
Figure 20 Cortical mRNA expression of APP processing genes. ................................. 120
Figure 21 Hippocampal sAPPα and Swedish sAPPβ levels. ....................................... 121
xvi
Figure 22 Cortical mRNA expression of Aβ-degrading enzymes. ................................ 122
Figure 23 Cortical mRNA expression of glial activation markers. ................................ 123
Figure 24 Cortical mRNA expression of PPAR-γ and LXR target genes. .................... 124
Figure 25 Cortical mRNA expression of RAS enzymes and receptors. ....................... 125
Figure 26 APP Processing Pathways. .......................................................................... 129
Figure 27 FMO boundaries from enriched progenitor cells. ......................................... 154
Figure 28 Morphology of Colony-Forming Assays. ...................................................... 155
Figure 29 SA-β-Gal Staining of CFU-Hill Assay. .......................................................... 156
Figure 30 Flow Cytometry of AcLDL Uptake in Putative EPCs. ................................... 157
Figure 31 AcLDL uptake required 3–4 hours of incubation. ......................................... 158
Figure 32 CFU-Hill colony number did not correlate with AcLDL uptake. .................... 158
Figure 33 Circulating CD34+CD133+ cells significantly correlated with CFU-Hill colony
number. ................................................................................................................. 160
Figure 34 CD34
bright
population in CD34+CD133+CD309+ subsets. ........................... 162
Figure 35 RAS receptor expression on cell subsets. ................................................... 163
Figure 36 RAS receptor expression on putative EPC subsets. .................................... 164
Figure 37 Decreased Angiogenic Cell Subsets in MCI. ............................................... 189
Figure 38 Increased Plasma ICAM-1 in MCI. ............................................................... 190
Figure 39 Increased Plasma MIP-1α in MCI. ............................................................... 190
Figure 40 Flow Cytometric Subsets in MCI. ................................................................. 192
Figure 41 CFU-Hill Colonies in MCI. ............................................................................ 194
Figure 42 SA-β-Gal Activity in MCI. ............................................................................. 198
xvii
List of Tables
Table 1 Currently approved drugs for dementia of AD type. .......................................... 18
Table 2 Central angiotensin receptors, ligands, and major responses in the CNS. ....... 49
Table 3 Mice exhibited no side preferences on the T-maze. .......................................... 93
Table 4 PCR Primer Sequences. ................................................................................. 118
Table 5 Panel of fluorescence-minus-one (FMO) controls. .......................................... 146
Table 6 Correlations between CFU-Hill colonies and flow cytometric cell subsets. ..... 159
Table 7 Correlation between CFU-Hill colonies and Framingham risk scores. ............ 161
Table 8 Correlation between CFU-Hill colonies and hematological parameters. ......... 161
Table 9 RAS receptor expression on putative EPC subsets. ....................................... 164
Table 10 Supplemental Neuropsychological Measures. .............................................. 176
Table 11 Plasma Biomarkers Measured by Immunoassay .......................................... 179
Table 12 Participant Characteristics. ............................................................................ 187
Table 13 Flow Cytometry Subsets in Cognition. .......................................................... 193
Table 14 CFU-Hill Colonies in Cognition. ..................................................................... 194
Table 15 Angiogenic cell subsets positively correlated to tests of verbal and visual
memory. ................................................................................................................ 196
xviii
List of Abbreviations
A(1-7) angiotensin-(1–7)
3xTg-AD triple transgenic mouse model
ACE angiotensin-converting enzyme
ACE2 angiotensin-converting enzyme 2
ACh acetylcholine
AChE acetylcholinesterase
AcLDL acetylated low density lipoprotein
AD Alzheimer's disease
Ang II angiotensin II
Ang-(1–7) angiotensin-(1–7)
AP aminopeptidase
APOE apolipoprotein E
APP amyloid precursor protein
ARB angiotensin II receptor blocker
AT1 angiotensin II receptor type 1
AT2 angiotensin II receptor type 2
Aβ amyloid-beta
BBB blood-brain barrier
BM bone marrow
CAC circulating angiogenic cell
CAND candesartan
CBF cerebral blood flow
CD cluster of differentiation
CDR Clinical Dementia Rating
CFU colony-forming unit
CNS central nervous system
CSF cerebrospinal fluid
CVO circumventricular organ
ECE endothelin-converting enzyme
EPC endothelial progenitor cell
FC flow cytometry
G-CSF granulocyte-colony stimulating factor
GM-CSF granulocyte macrophage-colony stimulating factor
HPC hematopoietic progenitor cell
HSC hematopoietic stem cell
IDE insulin-degrading enzyme
IL interleukin
xix
LTP long-term potentiation
MCI mild cognitive impairment
MMSE Mini-Mental State Examination
MRI magnetic resonance imaging
NACC-UDS National Alzheimer's Coordinating Center Uniform Data Set
NEI novel exploration index
NEP neprilysin
NonTg non-transgenic
NOR novel object recognition
PBMC peripheral blood mononuclear cell
PCR polymerase chain reaction
PPAR-γ peroxisome proliferator-activated receptor gamma
PPI prepulse inhibition
PS1, PSEN1 presenilin 1
RAS renin-angiotensin system
SA-β-gal senescence-associated β-galactosidase
SCO scopolamine
VaSC Vascular Senescence and Cognition
VaD vascular dementia

xx
Abstract
Introduction: Alzheimer's disease (AD) is an insidious neurodegenerative disease that
initially manifests as memory loss. Cognitive function progressively worsens over time,
and ultimately results in death. Modifiable risk factors such as midlife hypertension can
increase the risk of developing AD later in life. Clinical studies have found that
angiotensin II receptor blockers (ARBs) may confer additional protection against
cognitive decline beyond its antihypertensive actions. The mechanisms underlying this
additional protection may lie beyond controlling blood pressure, but rather through the
metabolism of angiotensin II to angiotensin-(1–7) [Ang-(1–7)]. Patients with AD also
have reduced numbers of functional circulating angiogenic cells, perhaps reflecting
diminished progenitor reserves. However, the role of these angiogenic cells early in
cognitive decline is unknown. This dissertation consists of two main parts: (1) a
preclinical study investigating the efficacy of the ARB candesartan and Ang-(1–7) in an
animal model of AD and (2) an observational clinical study of circulating angiogenic
cells in older adults with mild cognitive impairment (MCI).

Methods: In the preclinical study, triple transgenic AD (3xTg-AD) mice were treated
with candesartan, Ang-(1–7), or both for 8 months. Hippocampal amyloid-beta (Aβ) was
assessed by immunoassay, and potential underlying mechanisms were investigated
including: endothelial progenitor cells (EPCs), carotid blood flow, and Aβ clearance
pathways. In the clinical study, adults (55 years of age or older) underwent a
xxi
neuropsychological exam, MRI brain scan, and blood draw. Circulating angiogenic cells
were characterized by flow cytometry and colony-forming assay.

Results: 3xTg-AD mice treated with candesartan and Ang-(1–7) showed significant
reductions in hippocampal insoluble Aβ. Candesartan increased circulating EPCs,
reduced carotid blood flow, and increased the mRNA expression of Aβ-degrading
enzymes insulin-degrading enzyme (IDE) and endothelin-converting enzyme 2 (ECE2),
and microglial activation markers CD45 and Iba1. Participants with MCI exhibited
significantly reduced progenitor cell subsets CD34 and CD133. Angiogenic cell subsets
positively correlated with performance on verbal and visual memory tests. Voxel-wise
regression analyses revealed that participants with higher CD34 cells correlated to
increased volumes in the lingual gyrus, posterior cingulate gyrus, and precuneus.

Conclusion: Long-term treatment with candesartan and Ang-(1–7) significantly reduced
hippocampal Aβ in a mouse model of AD, which may be attributed to increased Aβ
clearance potentially through the increased expression of Aβ-degrading enzymes IDE
and ECE2. Circulating progenitor cells were found to be decreased in MCI, and
positively correlated to memory tests and brain regions involved in verbal and visual
memory. These results suggest that circulating progenitor cells may have a protective
role in the early stages of cognitive decline, which may serve as a useful diagnostic,
biomarker, or point of intervention.

1
Chapter 1: Introduction & Background
Alzheimer's Disease
Alzheimer’s disease (AD) is the most common form of dementia, which affects 5.3
million Americans. Patients experience cognitive decline, possibly due to amyloid
deposition and neurofibrillary tangles in the brain. Despite advances in our
understanding of the disease, no approved treatments can slow the progression of AD.
Pathology
Dr. Alois Alzheimer first described the disease in 1901, noting a patient with a
remarkably early-onset form of dementia. The 51-year-old female patient was admitted
with symptoms of short-term memory deficits, delusions, and hysteria. Upon her death
in 1906, Dr. Alzheimer commissioned an autopsy of her brain. He identified the
presence of amyloid plaques and neurofibrillary tangles, the key neuropathological
factors that would later be known as AD (Maurer et al., 1997).
The neuropathological assessment of AD involves the evaluation of AD-specific
neuropathological changes and their correlation to clinical and neuropsychological
exams. The characteristic neuropathological changes in AD include the presence of
neuritic plaques, extracellular amyloid-beta (Aβ) deposits, and neurofibrillary tangles
(NFTs) composed of phosphorylated tau (p-tau). Braak et al. developed a scoring
system to define the neuropathological staging of AD depending on the distribution and
density of these neuropathological markers at autopsy (Braak and Braak, 1991; Braak
et al., 2006). NFTs exhibited a consistent, characteristic distribution pattern that can be
classified by six distinct stages. These stages depict increasing tangle burden that
2
progress from the transentorhinal region (Stages I–II), to further limbic regions (III–IV),
and finally to widespread isocortical regions (V–VI).
An alternative model of neuropathological staging of AD has been proposed by Jack
et al., which incorporates in vivo AD biomarkers, brain structure, and clinical function
(Jack et al., 2010). This model proposes that Aβ markers become elevated prior to
alterations in brain structure, memory, and clinical function. Incorporating these facets of
AD pathogenesis may provide clinical utility in identifying the stages of cognitive decline
in a holistic manner. Importantly, this model relates temporal changes in AD biomarkers
to clinical disease stages.

Figure 1 Jack model of AD pathological cascade.
Reprinted with permission from Elsevier (Jack et al., 2010).

3
Epidemiology
The risk of developing AD increases with age (Keene et al., 2016). Over 46 million
people live with dementia worldwide, and projections estimate this figure will rise to over
130 million people by 2050 (Prince, 2015).
Epidemiological studies estimate that 4.7 million Americans over the age of 65 years
lived with AD in 2011. Of these, 0.7 million people were between 65 and 74 years, 2.3
million were between 75 and 84 years, and 1.8 million were 85 years or older. As life
expectancy increases and the baby boomer generation enters old age, it is estimated
that this figure will rise to over 13.8 million Americans by 2050 (Hebert et al., 2013).
The incidence of dementia doubles every 10 years after the age of 65 years (Prince
et al., 2013). There is no sex-dependent difference in the incidence or prevalence of
dementia or AD, but in terms of absolute counts more women live with the disease due
to differences in life expectancy, as women tend to live longer than men.
Hypotheses of Alzheimer's Disease
The development of treatments for AD have been based on two major hypotheses:
(1) the Aβ cascade hypothesis and (2) the cholinergic hypothesis. These hypotheses
were mainly founded upon neuropathological findings in AD: the presence of Aβ
plaques (particularly in the cerebral cortex and hippocampus), a loss in cholinergic
neurons from the basal forebrain, and other features such as neurofibrillary tangles.
These hypotheses have driven therapeutic efforts to clear Aβ or restore cholinergic
function.
4
Aβ is a 36–43 amino acid peptide, produced by the cleavage of amyloid precursor
protein (APP), a gene encoded on chromosome 21. Aβ is constitutively produced
throughout life in the brain. The Aβ cascade hypothesis proposes that the accumulation
of amyloidogenic species such as Aβ1–42 causes neurodegeneration that leads to
clinical AD.
A facet of the Aβ cascade hypothesis points to certain genetic causes of Aβ
accumulation. Autosomal dominant mutations of the APP gene or APP-cleaving genes
presenilin-1 (PSEN1/PS1) or presenilin-2 (PSEN2/PS2) can result in an overproduction
of Aβ, leading to an early-onset familial form of AD. The identification of these genes
has spurred the development of transgenic mouse models of AD that recapitulate Aβ
pathology and behavioral abnormalities. Patients with Down syndrome, who have an
additional APP gene due to chromosome 21 trisomy, also develop AD pathology and
symptoms 10–20 years earlier than most cases of AD. However, the vast majority
(~95%) of common AD cases are late-onset with no certain genetic cause.
Aβ accumulation can be interpreted as a net balance between Aβ production and
clearance. A net-positive accumulation can lead to AD. In the amyloidogenic pathway,
APP is processed to Aβ species by the sequential enzymatic action of β- and γ-
secretases. Alternatively, APP can be processed to non-amyloidogenic forms through α-
secretase (Gorelick et al., 2011).
On the other hand, Aβ clearance likely occurs through multiple processes acting in
concert including cellular, molecular, and enzymatic pathways (Wang et al., 2006b).
These include glial phagocytosis, perivascular drainage, binding to chaperone proteins
5
including apolipoprotein E (APOE), apolipoprotein J and α2-macroglobulin (α2M), and
receptor-mediated transport into and out of the bloodstream through low density
lipoprotein receptor-related protein 1 (LRP1) and receptor for advanced glycation
endproducts (RAGE).
Several Aβ-degrading enzymes can act upon and degrade Aβ as a substrate
(Santos et al., 2011). Insulin-degrading enzyme (IDE) and neprilysin (NEP) are the two
major enzymes responsible for Aβ clearance. To a lesser degree, angiotensin-
converting enzyme (ACE) and endothelin-converting enzymes (ECEs) may also be
involved. These enzymes are widely distributed throughout the brain, present in
neurons, microglia, astrocytes, and endothelial cells, amongst other cell types. These
enzymes can reside on the cell surface, in subcellular compartments, and as secreted
soluble forms.
Both IDE and NEP can degrade Aβ monomers and fibrils, and NEP has the added
advantage of being able to break down plaques. Overexpression of IDE or NEP in
transgenic animals has been shown to reduce or prevent amyloid plaque pathology
(Leissring et al., 2003). Further, a study of brain tissue from AD patients showed
reduced NEP mRNA, protein levels, and activity compared to normal controls (Wang et
al., 2010a). IDE activity was unchanged but IDE mRNA was significantly upregulated in
AD patients. Thus, IDE and NEP represent major proteases involved in Aβ degradation,
and there may be a compensatory effect in response to increasing Aβ load.
One common criticism of the Aβ cascade hypothesis is that Aβ plaque pathology
does not necessarily coincide with cognitive deficits. In the Nun Study of Aging and AD,
6
a continuing longitudinal study of American Roman Catholic nuns, a significant number
of elderly subjects developed severe AD neuropathology without accompanying
cognitive deficit (SnowdonNun Study, 2003). Rather, linguistic density served to be a
significant predictor of risk in developing AD later in old age. This disparity could be
explained by the greater toxicity of soluble and diffusible oligomeric forms of Aβ rather
than Aβ plaques. Aβ plaques have also been postulated to be a protective mechanism
to sequester the more toxic soluble forms.
The cholinergic hypothesis of AD states that cognitive decline and pathology stem
from the reduced innervation of the cerebral cortex by cholinergic neurons. Cholinergic
neurons release the neurotransmitter acetylcholine (ACh). Anticholinergics were found
to affect memory; reduced levels of choline acetyltransferase, the enzyme responsible
for generating ACh, were observed in the brains of AD patients. AD patients also
demonstrated a significant loss in forebrain cholinergic neurons. The finding that AD
patients present with cholinergic deficit led to the development of anticholinesterase
treatments for AD, which inhibit the breakdown of ACh and prolong its activity within the
synapse. Acetylcholinesterase inhibitors remain the current standard line of therapy for
AD patients.
Clinical Presentation
The clinical presentation of cognitive decline is best described as an insidious
process (Wolk and Dickerson, 2017). Patients typically present with initial symptoms of
memory loss. A formal neuropsychological exam aids in further differential diagnosis.
Mild cognitive impairment represents the earliest detectable clinical phase of cognitive
7
decline, in which significant deficits present in at least one cognitive domain. Dementia
represents advanced cognitive decline, in which extensive cognitive impairment
significantly impedes an individual’s ability to live independently. AD and vascular
dementia (VaD) are the two most common forms of dementia. They differ in their clinical
presentation and etiology, and both can manifest together as mixed dementia.
Cognitive impairment is the most common and prominent feature of AD, particularly
memory impairment. Deficits in other cognitive domains may also coincide or occur after
memory impairment. Executive dysfunction and visuospatial impairments typically occur
early in the disease course, while language and behavioral symptoms typically occur
much later. Less commonly, patients may initially present with a primary complaint in
language, visuospatial, or executive function rather than memory (McKhann et al.,
2011). In these cases, memory deficits can be coaxed or elicited from AD patients at
presentation. Symptom progression in AD often proceeds insidiously, in a gradual,
subtle way.
Memory Impairments
There is clear progression and distinction in the type of memory impairment in AD
(Markowitsch and Staniloiu, 2012). Declarative episodic memory, or memory of events
at a time and place, is affected most prominently in AD. Episodic memory depends on
the medial temporal lobe structures, such as the hippocampus and entorhinal cortex.
Procedural memory and motor learning are affected later in the disease, and mainly
relies on subcortical structures. Semantic memory, which includes facts, vocabulary,
8
and concepts, is impaired later in AD. This type of memory depends on the neocortical
temporal regions, particularly the anterior temporal lobe.
Within episodic memory, there exists immediate recall, recent memory, and remote
memory. Recent memory is impaired early in AD, which depends on the hippocampus,
entorhinal cortex, and other supporting medial temporal lobe structures. In contrast,
immediate memory (encoded in the sensory association and prefrontal cortices) and
remote memory (memory lasting years) do not get affected until later in the disease.
Therefore, hippocampal-dependent impairments in recent memory appear early in AD,
while immediate recall and remote memory are initially spared.
Memory is assessed in patients during neuropsychological exams. A basic memory
test asks patients to learn and recall a series of words or objects immediately, and then
after a delay on the order of minutes. The ability to recall or recognize items, with or
without selective cues, may indicate the severity of memory deficit or whether the
memory deficit is related to encoding or retrieval, and thus specific to AD. Other memory
tests may ask patients about their orientation (where or when they are present), or
about current events.
Executive Dysfunction
Impairment in executive function is another major symptom of AD, and may present
early on in AD to varying degrees. Common symptoms of executive dysfunction include
diminished motivation, organization, and decision-making ability. Executive function
deficits are typically reported by family members or colleagues, and may be interpreted
as a lack of motivation, the inability to multitask, or the inability to complete tasks.
9
Reduced insight into deficits is also another common feature of executive
dysfunction, in which patients fail to recognize the extent of their deficits. As such, it is
common for patients to underestimate or underreport symptoms. Interviewing a person
familiar with the patient is often helpful and necessary, as it may provide insight into
cognitive impairment that the patient may be unaware of.
Loss of insight develops alongside overall disease severity (McDaniel and Edland,
1995), which may be associated with and exacerbate neuropsychiatric symptoms. AD
patients with preserved insight are more likely to be depressed, while those with
impaired insight are more prone to agitation, disinhibition, and psychosis (Harwood and
Sultzer, 2000; Mizrahi et al., 2006).
Clinical Evaluation
The clinical criteria for AD is based upon a history of progressive cognitive decline,
exclusion of other etiologies, and cognitive impairment in one or more domains (Larson,
2016b). A detailed clinical assessment provides the tools to make a rational diagnosis of
AD in most patients, although sensitivity and specificity are limited (McKhann et al.,
2011). Clinical evaluation typically initiates with neuropsychological testing and medical
history, and may be further supported by neuroimaging studies. Biomarkers may be
incrementally helpful or supportive in the diagnosis of AD or other neurological
diseases, but they are generally not recommended for routine diagnosis. Many
clinicians use standardized mental scales to screen and monitor the progression of
clinical dementia.

10
Neuropsychological Testing
Formal neuropsychological assessment can help evaluate individuals with cognitive
impairment and dementia. Care must be taken to administer cognitive testing under
standardized conditions with demographically appropriate norms. Neuropsychological
testing can be helpful in a number ways, including: (1) establishing a baseline for follow-
up over time, (2) distinguishing amongst dementias based on their differential
neuropsychological profiles, and (3) ascertaining competencies to guide lifestyle
recommendations for patients (e.g. driving safety, financial responsibilities).
Measuring Clinical Disease Progression
The clinical progression of dementia is typically measured using mental scales,
including the Mini-Mental State Examination (MMSE), the Montreal Cognitive
Assessment (MoCA), and the Clinical Dementia Rating scale. Although these scales are
not linear, studies have shown that AD patients decline by an average of 3–3.5 points
on the MMSE scale each year. A small proportion (< 10%) of patients exhibit a more
rapid decline of 5–6 points on the annual MMSE exam.
There is a litany of mental scales, all with their own advantages and disadvantages.
For example, the MMSE is widely used to assess cognition that can be quickly
administered in 10–15 minutes. However, it is limited to cognition, culturally biased
towards English speakers, and copyright issues have stymied its widespread adoption.
Ultimately, mental scales are screening tools that can be used to assess clinical course
of dementias.
11
The NACC-UDS (National Alzheimer’s Coordinating Center Uniform Data Set)
attempts to harmonize these different mental scales for uniformity of data in clinical trials
in the US. Currently many different scales are used, depending on the clinical trial.
The mean survival after diagnosis of AD ranges from 3–20 years, with an average
life expectancy of 8–10 years, depending on the degree of impairment at the time of
diagnosis. AD patients typically die from end-stage complications that relate to
advanced debilitation, such as dehydration, malnutrition, and infection.
Neuroimaging
In the clinical setting, neuroimaging is often administered as an adjunct to a
comprehensive exam for dementia, which includes a neuropsychological exam, medical
and prescription history, tissue biopsies, and other functional exams (Relkin, 2015).
Neuroimaging provides valuable information about the brain that cannot be otherwise
obtained, including morphological structure, metabolic activity, and Aβ presence.
Neuroimaging may be used to aid in the primary diagnosis of dementia, the presence of
comorbidities, or the diagnoses of other neurological disorders. Neuroimaging studies
may also be used as primary or secondary outcomes in clinical trials for AD.
Structural, functional, and molecular imaging techniques have elucidated the
progression of cognitive decline in AD, from normal cognition to MCI and dementia.
These techniques have shed light on preclinical manifestations of the disease, in which
patients do not yet display overt cognitive or behavioral deficits.
The gold standard for diagnosing AD is post-mortem brain analysis for the putative
neuropathological features of amyloid plaques, neurofibrillary tangles, and neuronal
12
loss. Structural abnormalities that accompany neuronal loss include cortical and
hippocampal atrophy, and ventricular enlargement. Neuroimaging studies allows
clinicians and researchers to peer into the brain to investigate possible diagnoses and
interventions across the spectrum of cognitive decline.
Structural Neuroimaging
Structural neuroimaging reveals physical details of the brain. The two most prevalent
structural imaging techniques for evaluating dementia are computed tomography (CT)
and magnetic resonance imaging (MRI).
CT creates 3D representations of the brain by imaging X-rays in serial sections. CT
scans can produce images with sub-millimeter resolution with only a few minutes of
scan time, but provide poor distinction between gray and white matter.
MRI utilizes strong magnetic fields at varying frequencies to cause atoms to
resonate and emit radio waves, a principle known as nuclear magnetic resonance
(NMR). In biomedical imaging, MRI can derive images from the NMR of hydrogen
atoms, which are abundant in water and fat.
While MRI can produce images with similar spatial resolution to CT, MRI provides
superior contrast. MRI scans can take 30–60 minutes, depending on the regions of
interest and the level of desired contrast. Unlike CT, MRI does not rely on ionizing
radiation, and thus may be a safer imaging technique. However, MRI is contraindicated
in claustrophobic individuals and those that have magnetic metal implants such as
pacemakers. When available and not contraindicated in patients, MRI is the preferred
imaging technique due to its superior contrast and lack of ionizing radiation.
13
Structural MRI Findings in Cognitive Decline
Structural MRI findings in AD may indicate generalized or focal atrophy, and white
matter lesions. Generalized findings refer to overall brain structure while focal findings
include those localized to a region, such as the hippocampus. Focal hippocampal
atrophy is one of the earliest structural findings in AD, along with other medial temporal
lobe structures such as the entorhinal cortex.
Pathological changes in AD occur in the cortical gray matter, characterized by Aβ
plaques and neurofibrillary tangles that lead to neuronal and synaptic loss. Taken
altogether, these cellular changes manifest as cerebral atrophy in specific brain regions.
A wealth of structural brain imaging studies have identified that the entorhinal cortex and
hippocampus are amongst the first sites affected by MCI, with further progressive
atrophy in AD (Yin et al., 2013).
Discrete volumetric measures vary based on age, sex, and other demographics. For
example, cerebral atrophy naturally occurs with aging, and AD exacerbates this
process. Studies estimate that the brain loses 0.5% of its volume each year in normal
aging, 1–2% in MCI, and 2–4% in AD (Frisoni et al., 2010). The AD Neuroimaging
Initiative (ADNI) has made age-corrected norms for hippocampal volumetry available,
which can predict conversion rates from MCI to AD in some cases (McEvoy et al.,
2011). Therefore, age and other factors must be included as covariates for statistical
analyses of structural markers.
Similarly, the hippocampal sulcus (or hippocampal fissure) refers to the non-
parenchymal space between the dentate gyrus and the subiculum. As the hippocampus
14
or other medial temporal lobe structures shrink, the sulcus increases in volume.
Enlargement of the hippocampal sulcus has been associated with medial temporal lobe
atrophy in AD (Bastos-Leite et al., 2006). Thus, the hippocampal sulcus can be inferred
as a surrogate measure for medial temporal lobe atrophy.
While ventricular enlargement is widely present in post-mortem AD brains, its role in
neuroimaging findings is more complicated. Essentially, it is difficult to decipher whether
ventricular enlargement occurs due to cerebral atrophy, or is a separate, distinct
phenomenon.
In regards to VaD, some diagnostic criteria including the NINDS-AIREN criteria
necessitate neuroimaging. VaD is associated with cortical and subcortical infarctions
and periventricular white matter lesions. White matter ischemic changes have been
implicated in age-related cognitive decline.
White matter lesions (WML), or white matter hyperintensities, occur as nonspecific
radiologic findings on MRI brain scans. Radiological WML are not confined to specific
neuropathology, and are prevalent in both normal aging and demented elderly
individuals. Significant associations have been found with age and hypertension,
ischemic stroke and silent/clinical infarctions, and other physiological phenomena.
Studies have found inconsistent relationships between WML and dementia.
Functional and Molecular Neuroimaging
Functional neuroimaging measures parameters related to neuronal or vascular
activity, such as cerebral perfusion or metabolic activity. Positron emission tomography
(PET) and single photo emission computed tomography (SPECT) are the two most
15
common functional imaging techniques for evaluating dementia. They both involve
injecting radiotracers intravenously to detect cerebral perfusion or glucose metabolism.
Cerebral perfusion is decreased in both AD and VaD; in AD, it relates to decreased
metabolic demand due to neuronal loss; in VaD, it relates to cerebral ischemia. Glucose
metabolism in the brain can be measured with the fluorodeoxyglucose (FDG)
radiotracer, an analog of glucose, on FDG-PET scans.
In animals, cerebral perfusion is typically measured by laser Doppler flowmetry. A
surrogate measure can be used, such as carotid blood flow, which measures flow in the
two major common carotid arteries that supply oxygenated blood to the brain. Carotid
blood flow can be measured by Doppler ultrasound.
Amyloid PET imaging is the most common molecular neuroimaging technique. It
measures Aβ load in the brain by injecting radiotracers that bind with high affinity to
fibrillar Aβ. While a select few of these tracers are approved in the US and the
European Union for radiological use, they are not routinely used for diagnosis of AD.
Issues with how much ligand binding constitutes a positive amyloid result limits its use.
Positive amyloid findings also do not support an AD diagnosis with great confidence or
rule out other co-existing pathologies.
Biomarkers
Efforts have been made to identify minimally invasive biomarkers for detecting AD,
such as blood proteins. These dysregulated signaling proteins in AD are being further
investigated in those with MCI and pre-symptomatic individuals.
16
These observational studies for biomarkers in AD have led to other therapeutic
avenues. For example, deficits in certain signaling proteins such as hematopoietic
growth factors, granulocyte-colony stimulating factor (G-CSF) and granulocyte
macrophage colony-stimulating factor (GM-CSF), have led to utilizing clinically
approved recombinant forms to counteract these deficits.
There has been an ongoing interest in identifying potential biomarkers that can
correlate to AD. Plasma Aβ40 and Aβ42 can be measured in peripheral blood (Irizarry,
2004), but do not contribute to identifying AD with good reproducibility, and likely do not
reflect the Aβ processing in the brain (Vanderstichele et al., 2000). In contrast, some
cerebrospinal (CSF) biomarkers have been shown to reliably correlate to AD, but
require invasive lumbar punctures.
As an easily accessible tissue biopsy, plasma biomarkers are an attractive
diagnostic biomarker. A plasma biomarker can aid clinicians in evaluating patients with
probable AD, as well as identify possible individuals with pre-symptomatic AD for early
intervention. Although most proteins evaluated in serum or plasma may be useful as
biological correlates of AD, in that they are statistically different in AD patients compared
to normal controls, one of the major problems is that they show poor sensitivity and
specificity necessary for diagnosis (Irizarry, 2004).
An analysis of 120 signaling proteins in plasma revealed 18 potential biomarkers
that can reliably distinguish healthy controls from AD patients with close to 90%
accuracy (Britschgi and Wyss-Coray, 2009; O’Bryant, 2010; Ray et al., 2007).
Subsequently, these potential biomarkers were then studied in 47 patients with MCI and
17
assessed for their utility in identifying MCI patients that convert to clinical AD. These
proteins were found to be dysregulated in 20 of the 22 identified MCI patients that
converted to AD, while they remained within normal levels in patients with non-AD
dementia. Systems analysis of the 18 identified proteins revealed dysregulation in the
hematopoiesis, immune response, and apoptosis.
Further studies attempted to validate previously discovered plasma biomarkers
associated with AD with neuroimaging measures (Hye et al., 2014). It is possible that no
singular biomarker can reliably predict progression to dementia. Rather, biomarker
signatures composed of a combination of several biomarkers may be needed instead.
As such, the Penn Biomarker Core in the ADNI has recognized this and is developing a
composite of likely candidates that would suit a CSF and plasma biomarker signature
(Trojanowski et al., 2010). Initial CSF studies have revealed that a combination of CSF
Aβ42 and total tau effectively delineated the boundary between MCI and mild AD. A
temporal data model based on these biomarkers was generated, suggesting that Aβ
markers become abnormal first, followed by changes in neurodegenerative markers
such as CSF total tau, FDG-PET, and MRI that coincide with clinical symptoms
(Trojanowski et al., 2010). Evidence from these groups and others suggest that blood
protein signatures may be more helpful in successfully diagnosing AD compared to a
single biomarker alone (Lista et al., 2013).
Current Therapies
Current therapies for dementia provide a modicum of symptomatic relief, but do not
slow disease progression. There is no cure or disease-modifying therapy for AD.
18
Besides minimizing vascular risk or other modifiable factors, treatment with
cholinesterase inhibitors or in combination with an NMDA antagonist remains the
standard line of therapy. Physicians typically prescribe cholinesterase inhibitors for mild-
to-moderate AD (mild MMSE 21–26, moderate MMSE 10–20). As the disease
progresses to advanced dementia (MMSE < 17), combination therapy with memantine
is advised.
Behavioral disturbances such as agitation, aggression, apathy, delusions,
depression, disinhibition, hallucinations, and sleeping problems may be managed by
neuropsychiatric treatment.
The following table lists the current therapies approved for AD.

Drug Name Brand Name Approved In Class
tacrine Cognex 1993 cholinesterase
inhibitor
donepezil Aricept 1996
cholinesterase
inhibitor
rivastigimine Exelon 2000
cholinesterase
inhibitor
galantamine Razadyne 2001
cholinesterase
inhibitor
memantine Namenda 2003 NDMA antagonist

Table 1 Currently approved drugs for dementia of AD type.

19
Cholinesterase Inhibitors
AD patients have reduced cerebral levels of choline acetyltransferase, the enzyme
responsible for generating ACh, a key neurotransmitter in the brain. ACh deficit leads to
reduced cholinergic function in the brain. ACh esterase (AChE) inhibitors work by
blocking the breakdown of ACh, thereby allowing ACh to persist longer in the synapse.
Four AChE inhibitors are approved by the United States (US) Food and Drug
Administration (FDA) for treating dementia of the AD-type: tacrine, donepezil,
rivastigmine, and galantamine. Physicians no longer prescribe tacrine due to concerns
with hepatic toxicity and severe gastrointestinal side effects.
The most common side effect of AChE inhibitors is GI upset (diarrhea, nausea, and
vomiting), which is not particularly surprising due to the abundance of ACh receptors in
the GI tract. Alternative patch formulations, such as those for rivastigimine, may be
beneficial for patients that experience GI upset, since the transdermal route bypasses
gastrointestinal metabolism.
NMDA Antagonists
Memantine is an N-methyl-D-aspartate (NDMA) antagonist indicated for the
treatment of moderate-to-severe AD. NDMA antagonists work by blocking the actions of
NDMA receptors. Glutamate, the principal excitatory neurotransmitter in the brain, binds
to NDMA receptors and causes calcium influx and subsequent membrane
depolarization. Persistent glutamatergic activation of NDMA receptors has been
observed in AD. Excessive NDMA stimulation can lead to excitotoxicity and
20
neurodegeneration. Memantine has been proposed to be neuroprotective due to this
mechanism of action, but this has yet to be proven.
Risk Factors for Cognitive Decline & Dementia
An all-too-common theme has been identified in the research of AD. Observational
studies have identified risk factors or dysregulated metabolism in AD, which have led to
therapeutic efforts to reduce risk or compensate/correct for metabolic dysregulation.
However, most of these efforts have led to no significant therapeutic intervention.
Despite the lack of significant therapeutic intervention, some of these observations
have paved the way for possible preventive strategies in mitigating cognitive decline
and dementia. Risk factors for developing dementia include: age, genetics, cognitive
status, and vascular risk burden (Larson, 2016a).
Age
Age is the strongest risk factor for dementia, especially of the AD type. The
incidence of dementia doubles every 10 years after the age of 60 years (Prince et al.,
2013). As a result, the prevalence of dementia exceeds 50% in individuals over the age
of 90 years.
Genetics
Genetic factors are known to play a role in AD, particularly the early-onset cases.
While they only account for less than 1% of cases of AD, early-onset genetic factors
follow an autosomal dominant inheritance pattern related to mutations that alter amyloid
processing. Mutations in APP, PSEN1, or PSEN2 have been associated with an
increased risk of developing early-onset AD.
21
Other genetic factors increase the susceptibility of getting AD later in life, such as
apolipoprotein E (APOE). APOE plays a major role in cholesterol transport. Variations in
the APOE gene have been associated with AD. The ApoE-ε4 allele confers an 11 times
greater risk of developing AD.
Mild Cognitive Impairment
Mild cognitive impairment (MCI) is the intermediate stage between normal cognition
and dementia. In this stage, a patient may present with cognitive deficit in one or more
domains. The degree of cognitive deficit in MCI is not severe enough to prevent the
patient from living self-sufficiently, which does not satisfy the requirement for dementia.
The ability to diagnose patients at an earlier stage of cognitive decline, such as MCI, is
evidently growing increasingly important as many interventional clinical trials have failed
at the dementia stage. Older adults with MCI have an increased risk of developing AD;
patients with MCI convert to dementia at a rate of 10% per year. There has been
increasing interest in identifying the factors involved in the conversion of MCI to
dementia.
Vascular Contributions to Cognitive Impairment and Dementia
Vascular risk factors have been associated with an increased risk of age-related
cognitive decline and dementia, due to both AD and VaD independently (Gorelick et al.,
2011). The strongest association was found in midlife cardiovascular risk factors rather
than late life. Midlife cardiovascular risk factors appear to increase the rate of cognitive
decline in late life. These significant findings, while temporally separate, suggest that
22
preventive intervention in midlife may stave off cognitive decline. Unlike age and
genetics, vascular risk factors may be reversible and modifiable.
Hypertension
Midlife hypertension correlates with an increased risk of AD later in life (Launer et al.,
2000; Posner et al., 2002; Skoog et al., 1996; Yoshitake et al., 1995). A study following
13,000 middle-aged adults (48–67 years of age) for 20 years found that those with
uncontrolled hypertension at the start of the study performed 6.5% worse on cognitive
tests than their controls (Gottesman et al., 2014). This equated to about 2.5 years of
additional cognitive aging. Participants taking antihypertensive medications performed
better on cognitive tests than people with uncontrolled hypertension. Their results
suggest that controlling blood pressure in midlife protects against cognitive decline later
in life. Recent evidence suggests that angiotensin receptor blockers (ARBs) may confer
additional cognitive protection compared to other antihypertensive drug classes, a topic
discussed later in this chapter.
Hypertension exacerbates cognitive impairment, particularly affecting executive
function (Grigsby et al., 2002; Royall et al., 2004; Vicario et al., 2005). The renin-
angiotensin system (RAS) is a peptide hormone system that regulates blood pressure
and fluid balance. It has also been shown to be involved in regulating and maintaining
cerebral blood flow. Angiotensin II (Ang II) has been shown to decrease cerebral blood
flow and impair neurovascular coupling (Kazama et al., 2004; Saavedra and Nishimura,
1999).

23
Cerebrovascular Reactivity
The brain depends on a continuous supply of blood that changes in response to
energy requirements (Hossmann, 1994). In activated brain regions, cerebral blood flow
(CBF) increases to meet increased energy demands and clear metabolites generated
by cellular respiration (Attwell and Iadecola, 2002). This phenomenon, known as
cerebrovascular reactivity or functional hyperemia, describes a change in cerebral blood
flow in response to a vasoactive stimulus (Fierstra et al., 2013). Both CBF and
cerebrovascular reactivity are impaired in the early stages of AD (Hock et al., 1997;
Jagust, 2000; la Torre, 2004; Mentis et al., 1998; Prohovnik et al., 1988).
Hypertension induces cerebrovascular remodeling mediated through Ang II
(Iadecola and Gorelick, 2004). This cerebrovascular remodeling can negatively impact
cerebrovascular reactivity by blunting the increase in CBF (Faraci and Heistad, 1998).
These alterations are postulated to underlie the cognitive impairment and brain damage
associated with hypertension (Droste et al., 2003; Iadecola and Gorelick, 2004).
Evidence suggests that impaired cerebrovascular reactivity coincides with and
exacerbates AD pathogenesis, which may be due in part to chronic Ang II activation of
its constitutive AT1 receptor.
Ang II has been shown to attenuate the CBF increase produced by activation of the
mouse somatosensory cortex (Kazama et al., 2003). The mouse somatosensory cortex
can be activated by mechanical stimulation of the whiskers, which consequently results
in an increase in CBF. This alteration in the neurovascular coupling between neural
24
activity and CBF suggests that Ang II-associated hypertension may be contributing to
further brain dysfunction.
Kazama et al. investigated whether the Ang II-induced attenuation in CBF increase
was mediated through the activation of AT1 receptors and production of reactive oxygen
species (ROS) through NADPH oxidase (Kazama et al., 2004). Coadministration of Ang
II with the ARB losartan or ROS scavengers, superoxide dismutase or tiron, showed no
attenuation in CBF increase. Further, the effects of Ang II attenuation on CBF increase
was not observed in mice lacking the gp91phox subunit of NADPH oxidase or in wild-
type mice treated with an NADPH oxidase inhibitor. Ang II did not increase ROS
production in the gp91-null mice. Taken altogether, these results demonstrate that Ang II
impairs cerebrovascular reactivity through AT1 activation and ROS production through
NADPH oxidase.
Interestingly, the effects of Ang II on neurovascular dysfunction may be dependent
on sex. In the same Ang II-whisker stimulation model of cerebrovascular reactivity, the
CBF increase was attenuated in only male but not female C57BL/6J mice (Girouard et
al., 2008). Females showed reduced susceptibility to the effects of Ang II. This effect
was abolished by the removal of the ovaries, and reinstated with estrogen
administration to the ovariectomized mice. Administration of estrogen in the male mice
also abolished the Ang II-induced attenuation in CBF increase. These results suggest
that female mice are less susceptible to the cerebrovascular dysregulation related to
Ang II, likely related to estrogen. As such, women may be protected from the
cerebrovascular complications resulting from hypertension.
25
Animal Models of AD
Modeling AD in animals is a complex task that involves carefully selecting the
appropriate model of disease in full recognition of their advantages and limitations.
Mouse models of AD have spurred the discovery and development of therapeutics and
diagnostics for AD (Chin, 2011). They differ depending on their genetic basis, length of
cognitive decline, and Aβ deposition amongst many other factors.
Transgenic mouse models of AD based on familial mutations are by far the most
popular model of dementia, and represent a starting point for research in aging and
cognitive decline. While cell based models of AD have utility for screening compounds,
they do not recapitulate the complexity of mammalian brain physiology and behavior.
It is important to note that rodents do not sporadically develop AD, due to the
differences in their APP sequence and propensity to aggregate. Human and mouse APP
differ by 17 amino acids, 3 of which reside in the Aβ sequence (amino acids 3, 10, 13).
Murine Aβ does not accumulate nor form extracellular deposits, even when
overexpressed (Jankowsky et al., 2007). Monoclonal antibodies have been developed
that bind to epitopes specific for human Aβ, such as 6E10 that reacts to the Aβ1–16
epitope. Antibodies such as 4G8 bind to both human and mouse Aβ, since they react to
Aβ17–24 common to both humans and mice.
The most common APP mutations associated with familial AD have been used in
transgenic animal models, including the Swedish double mutation (K670N and M671L),
London (V717I) or Indiana (V717F) mutations, Arctic mutation (E693G), and Dutch
mutation (E693Q). These mutations confer more favorable substrates for pathogenic
26
APP processing through β- and γ-secretase, or increase the ratio of aggregation-prone
species such as Aβ42/Aβ40. Familial AD mutations have spurred the development of
myriad mouse models of AD, covered more in depth by Chin (Chin, 2011). These
models attempt to replicate the neuropathological hallmarks of the disease, namely
amyloid plaques and neurofibrillary tangles.
The triple-transgenic mouse model (3xTg-AD) harbors APP
Swe
, PS1
M146V
, and
tau
P301L
transgenes. APP
Swe
and tau
P301L
transgenes under control of the mouse Thy1.2
regulatory element result in overexpression of mutant human APP and tau, while the
knock-in PS1
M146V
mutation confers increased catalytic activity by γ-secretase.
Consequently, the 3xTg-AD mouse model develops both characteristic pathologies
seen in AD: Aβ plaques and neurofibrillary tangles. Extracellular amyloid deposits first
develop in the cortex at 6 months of age, and then progress to plaque formation in the
hippocampus at 12 months of age. Shortly thereafter tangles appear in the
hippocampus.
Behavioral Tasks
The large majority of cognitive tasks in animal models of AD involve spatial maze
tasks (Buccafusco et al., 2009). There are also cued and recognition memory tests,
which have analogs to human cognitive tests. Successive behavioral tasks should be
administered from less aversive to most aversive, to reduce stress as a confounding
factor.
27
For many years, the gold standard has been the Morris Water Maze task. A less
aversive dry version has been adapted and is gaining popularity, called the radial arm
maze task. Another popular dry task is the T-maze (Deacon and Rawlins, 2006).
Recognition tests assess the ability of animals to recognize previously learned
objects or locations. The Novel Object Recognition (NOR) test involves familiarizing an
animal to discrete objects, introducing a new object, and assessing their preference to
the novel object. Rodents tend to prefer novel objects. The assumption is that rodents
with cognitive deficits have difficulty remembering the old object or location, and thus
would explore both familiar and novel objects with equal preference.
Other behavioral disturbances can be assessed, such as neuropsychiatric
anomalies like anxiety and depression. Anxiety-like effects can be assessed by a
rodent’s thigmotactic behavior (the tendency to prefer walls) on the open-field test or
performance on an elevated plus maze.
Deficits in sensorimotor gating have been observed in prominently in schizophrenia
and anxiety disorders (Braff et al., 2001; Swerdlow et al., 2008). Sensorimotor gating is
the ability to filter out important sensory information from background noise. Studies
have shown that cholinergic deficits have been associated with impaired sensorimotor
gating, although these sensorimotor impairments do not appear to be present in clinical
cases of amnestic MCI or mild AD (Hejl et al., 2004; UEKI et al., 2006).
The prepulse inhibition (PPI) task can assess visual and/or audio sensorimotor
gating in both humans and animals (Geyer and Dulawa, 2003; Geyer and Swerdlow,
1998). PPI tasks measure the motor response of a subject to a prepulse stimulus (e.g. a
28
loud noise or bright light), followed by another stimulus. Subjects with intact PPI elicit a
reduced motor response to the second stimulus. Transgenic APP/PS1 mice exhibit both
PPI and Morris Water Maze deficits at 7 and 22 months of age (Wang et al., 2012).
The disparity in PPI between animal models and clinical AD patients may be related
to APP or presenilin mutations, since the PPI deficits presented only in transgenic
animal models with familial AD mutations. On the other hand, the observations of PPI in
clinical cases of MCI or AD were not limited or confined to only individuals with rare
familial mutations. Also, the clinical cases observed no sensorimotor deficits in the early
stages of the cognitive decline, which may become more apparent later in advanced
stages of dementia when behavioral disturbances become more prevalent.
Renin-Angiotensin System
The classical RAS, as it has been traditionally described for over 100 years,
regulates blood pressure and fluid balance. The role of the classical RAS is best known
for its homeostatic functions in the circulatory system. Beyond the classical RAS, local
organ- and tissue-specific RASs have been identified and described. The brain RAS is
covered in depth in a later section.
The RAS is a vasoactive hormone system that consists of oligopeptide ligands,
enzymes, and endogenous receptors. Enzymes cleave the ligands further to distinct
metabolites, which have their own separate receptors. Most of the angiotensin-related
enzymes are metallopeptidases that depend on metal cofactors for activity.
Angiotensinogen is a 453-amino-acid-long protein produced constitutively in the liver
and released into circulation. Renin, produced in the kidneys, cleaves the large
29
angiotensinogen protein to angiotensin I (Ang I). ACE and chymase cleave Ang I to the
vasoactive peptide Ang II. Ang II can bind to its constitutive Ang II type 1 (AT1) receptor,
which induces vasoconstriction and ultimately leads to an increase in blood pressure.
These elements form the classical RAS, also known as the ACE/Ang II/AT1 pathway.
Alternatively, Ang II can further metabolize to angiotensin-(1–7) [Ang-(1–7)] by the
action of angiotensin-converting enzyme-2 (ACE2). Ang-(1–7) can also be indirectly
formed by a two-step process: cleavage of Ang I through ACE2 to angiotensin-(1–9)
[Ang-(1–9)], and subsequently cleavage by ACE to Ang-(1–7). It is important to note that
the indirect pathway is less catalytically efficient (RICE et al., 2004). Ang-(1–7) can bind
to its endogenous receptor Mas, which elicits vasodilatory and anti-inflammatory effects
(Santos et al., 2003). These components form the protective RAS, also known as the
ACE2/Ang-(1–7)/Mas pathway.
NEP, also known as membrane metallo-endopeptidase (MME) and CD10, can break
down both Ang I and Ang-(1–9) to Ang-(1–7).
30

Figure 2 Simplified angiotensin pathway.
Renin breaks down angiotensinogen to Ang I (not shown). Ang I is cleaved by ACE to
the vasoactive peptide Ang II. Ang II can then bind to its constitutive receptor AT1 or its
inducible receptor AT2. Alternatively, Ang II can be cleaved by ACE2 to Ang-(1–7),
which binds to the Mas receptor.
AT1 AT2 Mas
Asp Arg Val Tyr Ile His Pro Phe His Leu
Angiotensin I
Asp Arg Val Tyr Ile His Pro Phe
Angiotensin II
Asp Arg Val Tyr Ile His Pro Phe His
Angiotensin-(1–9)
Asp Arg Val Tyr Ile His Pro
Angiotensin-(1–7)
ACE2
ACE
ACE2
ACE
31

Figure 3 Pathways of Ang-(1–7) generation.
The catalytic efficiency of angiotensin enzymes ACE, ACE2, and NEP is denoted by the
k
cat
/K
m
for each reaction (RICE et al., 2004).

Angiotensin enzymes act as Aβ-degrading enzymes
Key enzymes in the angiotensin metabolic pathway can also act upon Aβ as a
substrate (Santos et al., 2011). These RAS-related Aβ-degrading enzymes include NEP,
ACE, and ACE2. NEP-knockout mice exhibit behavioral impairments and increased
cerebral Aβ deposition (Madani et al., 2006). ACE can convert Aβ42 to Aβ40 and further
cleave Aβ40 species (Santos et al., 2011). Overexpression of ACE in myelomonocytes
drastically reduced both soluble and insoluble Aβ42 species, and prevented cognitive
decline in a mouse model of AD (Bernstein et al., 2014). It has been recently discovered
that ACE2 can break down Aβ43 species to less neurotoxic species (Liu et al., 2014).

32
Vascular Homeostasis and RAS Therapy
Vascular homeostasis is maintained by the balance between the classical and
protective RAS pathways (Rabelo et al., 2011). Imbalance results in a loss of
homeostasis. In the case of RAS-mediated hypertension, an overabundance of Ang II
can lead to chronic hypertension and increase the risk of deleterious cardiovascular
effects and disease.
Ang II’s major role in blood pressure regulation has made it a therapeutic target for
treating hypertension. Drugs have been developed that target either the production or
action of Ang II. The two main RAS-modifying drugs are: ACE inhibitors & angiotensin II
receptor blockers (ARBs).
ACE inhibitors target the production of Ang II by inhibiting ACE. The drug class
consists of the -prils, including lisinopril, enalapril, captopril, etc. Lisinopril is the most
widely used ACE inhibitor.
ARBs block the action of Ang II by directly antagonizing the constitutive AT1
receptor. ARBs consist of the -sartans; the FDA has approved eight ARBs for
hypertension: azilsartan, candesartan, eprosartan, irbesartan, losartan, olmesartan,
telmisartan and valsartan. Their molecular differences impart differential profiles in:
pharmacokinetics/pharmacodynamics, lipophilicity, volume of distribution, bioavailability,
half-life, AT1 receptor affinity, PPAR-γ activity, and their ability to penetrate the blood-
brain barrier (BBB). Losartan is perhaps the most widely used ARB. Candesartan and
telmisartan are thought to cross the BBB based on their lipophilic profile and radioligand
binding studies (Michel et al., 2013).
33
Both ACE inhibitors and ARBs are used in monotherapy and combination therapy to
treat hypertension. Some guidelines point to using ACE inhibitors and ARBs as first-line
therapies except in certain demographics, such as black patients and elderly patients
(Krause et al., 2011; Mancia et al., 2009). Data indicate that blacks and elderly patients
may be less sensitive to ACE inhibitors and respond better to calcium channel blockers
or diuretics for treating hypertension (Materson et al., 1995). This differential response
to RAS therapy may be related to lower baseline plasma renin activity in both blacks
and elderly patients (Blaufox et al., 1992).
ARBs are clinically preferred in some indications, specifically in patients with type 2
diabetes mellitus with proteinuria and/or renal insufficiency (Mallat, 2012). Randomized
clinical trials have shown that ARBs delay the progression of nephropathy (Association,
2008). Evidence has also shown that ARBs may delay the development of diabetes,
and thus prevent cardiovascular events and other diabetic complications in high-risk
patients (Izzo and Zion, 2011).
Brain RAS
While the classical RAS has been well described as an endocrine system for over a
century in its role in blood pressure regulation and fluid balance, the existence of non-
classical paradigms such as the protective RAS and local paracrine tissue-specific
RASs has emerged. Surmounting evidence has pointed to a local and independent
brain RAS with the necessary functional elements: angiotensinogen, angiotensin
peptides, enzymes, and receptors (Wright and Harding, 2013). Further investigations
have implicated the brain RAS in neurodegenerative diseases such as AD and
34
Parkinson’s disease (PD) (Wright et al., 2013). Here I compare the RAS in vascular and
brain compartments, elaborate on the impact of the RAS on cognitive decline and
dementia, and weigh the contribution of RAS therapy on these cognitive outcomes.
Cardiovascular and Central Pressor Actions of Angiotensins
The principle function of the peripheral RAS is blood regulation control directly
through AT1-mediated vasoconstriction. Cardiovascular control can also be indirectly
mediated through central actions of the brain angiotensin peptides as well (Wright and
Harding, 2013). Intracerebral administration of Ang II exert pressor actions through
sympathetic innervation, vasopressin release, and inhibition of the baroreceptor reflex
(Fitzsimons, 1998). These actions can promote drinking and sodium intake behavior.
Central hypertensive actions of Ang II/AT1 can be counteracted by the vasodilatory
actions of Ang-(1–7)/Mas activation (Iyer et al., 1998).
Current evidence supports the existence of two primary brain angiotensinergic
pathways: (1) a forebrain pathway that integrates circumventricular organs (CVOs) with
the paraventricular (PVN), supraoptic (SON), and median preoptic nuclei, and (2) a
pathway that connects the hypothalamus and medulla, which includes the area
postrema (AP) and nucleus of the solitary tract (NTS) (Llorens-Cortes and Mendelsohn,
2002). The first pathway involving CVOs are thought to mediate the interaction between
central and peripheral RASs.
Neuroanatomical and Cellular Sources of Angiotensins
The brain can be divided by neuroanatomical region, function, or cell type. Cerebral
regions can be divided by lobe (frontal, parietal, occipital, temporal, limbic). In particular,
35
the medial temporal lobe that includes the hippocampus and entorhinal cortex is
exquisitely linked to cognitive function and is the earliest affected structure in AD.
Brain cells include: neurons, glial cells (e.g. astrocytes, microglia, oligodendrocytes),
vascular endothelial cells, and pericytes. All elements of the RAS are differentially
expressed in these neuroanatomical regions and brain cells. Our appreciation and
understanding of the brain RAS and its organization at the regional and cellular level
continues to evolve (Davisson, 2003).
Local Production of Angiotensin Elements
Angiotensinogen and most angiotensin metabolites such as Ang II do not cross the
BBB in appreciable amounts due to their large size. CVOs lack a BBB and may be
susceptible to effects from the peripheral RAS and may be a source of local Ang II.
The brain RAS supports the genesis of angiotensinogen and active brain
angiotensin ligands: Ang II, Ang III, Ang IV, Ang-(3–7), and Ang-(1–7) (de Kloet et al.,
2015; McKinley et al., 2003). These ligands can bind to the angiotensin receptors AT1,
AT2, AT4, and Mas that are widely distributed throughout the brain. These receptors
may be involved in cognitive processing and memory formation based on their proximity
to these associated brain regions, such as the hippocampus and amygdala.
36

Figure 4 Brain RAS and BBB Permeability.
Schematic view of AT1 receptors and RAS-permeable regions of the rat brain in the
sagittal plane. Dotted areas represent high densities of AT1 receptors. Vertical-striped
areas represent CVOs that lack a BBB, and thus exposed to the peripheral RAS
(McKinley et al., 2003). Reprinted with permission from Elsevier.

Angiotensinogen and Renin
Most brain regions express angiotensinogen, particularly the medulla and
hypothalamus (McKinley et al., 2003). Angiotensinogen is located in the brain
extracellular fluid and the CSF. While neurons may produce angiotensinogen, glial cells
produce the vast majority of it, particularly in astrocytes. One study investigated a
transgenic rat model with a GFAP promoter coupled to the expression of an antisense
construct targeting angiotensinogen mRNA. These rats exhibited 90% lower levels of
angiotensinogen and Ang I, which suggested that astrocytes are the major source of
angiotensinogen in the brain (Schinke et al., 1999).
37
Renin, produced in the kidneys, is chiefly responsible for cleaving angiotensinogen
to Ang I in the peripheral RAS. However, central renin-angiotensin production is
controversial and less understood. Evidence for angiotensin production in the brain has
existed for several decades, but the responsible enzyme was thought to be cathepsin D
(McKinley et al., 2003).
There have been reports identifying or refuting the presence of renin mRNA, protein,
activity, and colocalization to angiotensinogen (Davisson, 2003; Li et al., 2012; McKinley
et al., 2003). However, the consensus is that renin expression is at or below the limit of
detection in most assays. Renin levels are thought to be extremely low in the brain and
localized to neurons and astrocytes (de Kloet et al., 2015). Other enzymes may be
involved in angiotensinogen catabolism, such as cathepsins and tonin, but their
relevance in the brain RAS has yet to be fully resolved (Cardoso et al., 2010; Klickstein
et al., 1982; Lomez et al., 2002).
Newer evidence suggests that prorenin and its receptor prorenin receptor (PRR)
may be actively involved in producing Ang II. Prorenin, a precursor of renin, has
catalytic activity that depends on a conformation in which the active site is exposed.
Binding of prorenin to PRR activates prorenin enzymatic activity. PRR localizes
abundantly in neurons, and in microglia and astrocytes to a lesser degree (de Kloet et
al., 2015). Notably, prorenin is the dominant form of total renin (prorenin and renin) in
the brain; PRR is highly expressed in the brain; PRR levels increase in hypertensive
animal models; knockdown of PRR in the hypothalamus is associated with reduced Ang
II in the same region; and reduced PRR expression is associated with reduced blood
38
pressure in Ang II-dependent hypertension (Li et al., 2012). Thus, the prorenin-PRR
interaction may serve as a rate-limiting step for angiotensin peptide formation in the
brain, and further studies are warranted. The search for elusive renin-related activity in
the brain continues.
Enzyme Distribution
ACE
ACE is extensively expressed in the central nervous system (CNS), with very high
expression particularly in the CVOs. Due to the lack of the BBB in this area, it has been
postulated that high ACE expression in the CVOs may produce local Ang II from
circulating peripheral Ang I (McKinley et al., 2003).
Within the human hippocampus, ACE was highly concentrated in the molecular layer
of the dentate gyrus, whereas it was almost completely absent in the granular layer.
Low concentrations were also observed in the subiculum and entorhinal cortex. Within
the cerebrovascular endothelium, moderate-to-high ACE expression was found on the
surface, while lower levels were associated with the smaller vessels within the brain
(Chai, 2013).
Some significant species differences exist in regards to ACE expression. For
example, high ACE expression was found in the rat choroid plexus, a region in direct
contact with CSF, while human choroid plexus only showed low expression (Chai,
2013). One could speculate that this could account for a species-dependent source of
Ang II in the CSF.
39
The wide distribution of ACE in mammalian brains suggests that it has a role beyond
the production of Ang II, and may be involved in the metabolism of other neuropeptides
(Rogerson et al., 1995). Beyond Ang I, ACE can act upon other substrates such as Aβ,
specifically breaking down Aβ42 to Aβ40 and subsequently smaller peptides (Santos et
al., 2011; Wang et al., 2006b).
Although ACE is primarily a membrane-bound protein, it can also be released as its
soluble form into extracellular fluid, plasma, and CSF by membrane-bound secretases
or sheddases (HOOPER and Turner, 2003; Oppong and Hooper, 1993).
ACE2
Immunohistochemical staining for ACE2 depicted wide expression throughout the
mouse brain, with abundant expression in the motor cortex, caudate putamen, nucleus
tractus solitary, lateral reticular nuclei, nucleus amibuus, and area postrema (Doobay et
al., 2007). Double labeling with GFAP and MAP2, a neuronal marker, revealed that
ACE2 expression was colocalized to the neuronal cytoplasm and not astrocytes.
Morphology of the immunostained cells also predominantly appeared to resemble
neurons. Similar to ACE, ACE2 can also exist in both membrane-bound and soluble
forms (Barrett et al., 2012). The presence of cytoplasmic ACE2 suggests that a
significant amount of neuronal ACE2 exists in the soluble form.
The only definitive physiological role for ACE2 is the cleavage of Ang II to Ang-(1–7).
Transgenic mice with either overexpressed AT1 or angiotensinogen and renin showed
differential ACE2 expression compared to their control non-transgenic littermates,
suggesting that ACE2 serves a regulatory function in the RAS (Doobay et al., 2007).
40
The expression level and activity of brain ACE2 also appear to be affected by other RAS
components, such as ACE and AT1 receptors (Huijing Xia, 2008).
Like ACE, ACE2 can also act upon other substrates. The role and significance of
ACE2 in the metabolism of other neuropeptides is only beginning to emerge. A recent
study in 2014 by Liu et al. discovered that ACE2 can convert Aβ43, a particularly
neurotoxic species that may sow the initial seeds for amyloid aggregates, to Aβ42 (Liu
et al., 2014). A combination of ACE2 and ACE can sequentially convert Aβ43 to Aβ42 to
Aβ40 species. Further, they showed that ACE2 activity was decreased in AD patients
compared to controls. These results suggest that peripheral and central ACE2 may be
implicated in the pathogenesis of AD and amyloid deposition.
Aminopeptidases A and N

Figure 5 Alternative Ang II Processing by Aminopeptidases.
An alternate pathway beyond the classical ACE/Ang II/AT1 and protective
ACE2/Ang-(1–7)/Mas exists in the brain. Aminopeptidases, peptidases that cleave N-
terminal peptide bonds, drive the metabolism of Ang II (1–8) towards Ang III (2–8), Ang
AT
1
Ang II
1–8
AT
4
Ang III
2–8
Ang IV
3–8
A-(3–7)
AP-A
AP-N
CP-P
PO
41
IV (3–8), and Ang-(3–7). Ang IV and Ang-(3–7) interact with a putative AT4 receptor,
which has been theorized to either be insulin-regulated aminopeptidase (IRAP) or the
hepatocyte growth factor (HGF)/c-Met receptor (Wright and Harding, 2011). The Ang
IV/AT4 pathway is perhaps best described by Wright, Harding, et al.
Aminopeptidase-A (AP-A) metabolizes Ang II to Ang III. Ang III binds to and activates
AT1 with similar affinity to Ang II. Ang III-mediated AT1 activation exerts the same
pressor effects, which is blocked by losartan. Aminopeptidase-N (AP-N) breaks Ang III
down to Ang IV, a peptide that binds to the AT4 receptor. Ang IV can further metabolize
to Ang-(3–7) by carboxypeptidase P (CP-P) and prolyl oligopeptidase (PO) cleavage.
AP-A and AP-N exist in the rodent brain (Réaux et al., 1999a; 1999b; Zini et al.,
1996). AP-A, similar to AT1 receptors, localize to the CVOs (SFO and OVLT), PVN,
SON, NTS, and RVLM (Gao et al., 2014). High AP-A enzymatic activity and
immunolabeling present in the hypothalamic nuclei and the medulla (De Mota et al.,
2008; Zini et al., 1996). Moderate AP-A activity was noted in human medial temporal
lobe structures (De Mota et al., 2008).
Autoradiography studies depict a wide distribution of AP-N in the rat brain (Noble et
al., 2001). High expression of AP-N was noted in brain microvessels, meninges, choroid
plexus, pineal gland, paraventricular nucleus, and pituitary gland. Moderate-to-high
levels of AP-N were observed in the hippocampus. Studies have localized both AP-A
and AP-N to the plasma membrane of pericytes in cerebral microvessels (Healy and
Wilk, 1993; Kunz et al., 1994).
42
Ang III is thought to be a principal mediator of central pressor effects. Administration
of EC33, a specific inhibitor for AP-A, abolished the pressor response to centrally
administered Ang II, suggesting that Ang II may need to be converted to Ang III to exert
its central pressor effect (Réaux et al., 1999b). Pharmacological interventions have
been developed to target central Ang III-mediated hypertension (Gao et al., 2014).
Initially thought to be an inactive peptide, studies investigating Ang IV administration
or AT4 agonism have determined that they play a key role in memory facilitation and
cerebroprotection. Ang IV, and its analog Nle
1
-Ang IV, have been shown to facilitate
long-term potentiation (LTP), learning, and memory consolidation (Wright et al., 2013).
Ligand Distribution
Ang II is the most prevalent vasoactive angiotensin peptide in the brain, while Ang-
(1–7), Ang III, Ang IV, and other metabolites are present to a lesser degree. The
prevalence of membrane-bound peptidases suggest that most ligand production occurs
in the extracellular space, although evidence for intracellular staining of Ang II and Ang-
(1–7) does not preclude intracellular production (Krob et al., 1998; Robbins et al., 2010).
Receptor Distribution
AT1 and AT2 Receptors
In rodents, AT1 receptors abundantly localize to areas that regulate blood pressure
and mineral balance, such as the PVN and SFO, while AT2 receptors localize to limbic
and thalamic regions that can indirectly regulate cardiovascular function (de Kloet et al.,
2015; Wright and Harding, 2011).
43
The cellular localization of these receptors remains debatable. Immunohistochemical
and in vitro studies have characterized AT1 and AT2 receptors on both neuronal and
non-neuronal cells, but some have argued that most angiotensin receptor antibodies are
non-specific (Benicky et al., 2012; Hafko et al., 2013). In situ hybridization studies
suggest that neurons predominately express AT1 and AT2 receptors under normal
conditions, while glial expression is absent (de Kloet et al., 2015). However, these in situ
hybridization studies have not studied receptor expression in pathological models. Since
studies have shown that AT1 activation mediates central inflammation and oxidative
stress (Kazama et al., 2004), pathological states may show differentially expressed
angiotensin receptors on neurons and glia.
AT4 Receptor
The characterization and localization of the AT4 receptor has been mainly conducted
using competitive binding assays against Ang IV (Wright and Harding, 2011). Therefore,
the putative AT4 receptor is defined as the endogenous receptor for Ang IV. Other
angiotensin agonists and analogs have been discovered to bind to the AT4 receptor,
including Ang-(3–7), Ang IV analogs, and LVV-hemorphin 7 (a fragment of the
hemoglobin β-chain) (Chai et al., 2000; Wright and Harding, 2011).
AT4 receptors are widely distributed in the brain, across multiple species, with high
density in the: anterior pituitary, caudate putamen, cerebellum, cerebral cortex, lateral
geniculate body, globus pallidus, habenula, hippocampus, inferior olivary nucleus,
nucleus basalis of Meynert, periaqueductal gray, piriform cortex, superior colliculus,
thalamus, and ventral tegmental area (Chai et al., 2000; Wright and Harding, 2011).
44
Mas Receptor
Central Mas receptors have been mostly characterized in rodent species thus far. In
situ hybridization data indicate that Mas gene expression are also localized similarly in
mice and humans, particularly in the dentate gyrus of the hippocampus and the piriform
cortex (Freund et al., 2012). The localization of Mas receptors within the hippocampus,
the region of the brain responsible for working memory, suggests that Mas may have a
role in learning and memory. Rats express Mas receptors in the hippocampus,
amygdala, cortex, and hypoglossal nucleus and in cardiovascular-related areas such as
the NTS, CVLM, RVLM, inferior olive, PVN, and supraoptic nucleus (Becker et al.,
2007). Mas receptor immunostaining appears to predominately localize to neurons.

45

Figure 6 Overview of Brain RAS Pathways.
Brain RAS pathways involve multiple angiotensin peptides, enzymes, receptors, and
receptor antagonists. Carboxypeptidases drive the metabolism of Ang II towards Ang-
(1–7), while aminopeptidases drive the metabolism towards Ang III, Ang IV, and Ang-
(3–7).

AT
1
AT
2
Ang II
1–8
Mas
A-(1–7)
A-(1–9)
Ang I
1–10
ACE
ACE2
ACE
ACE2
ARB PD123,319 A-779
Angiotensinogen
Renin
AT
4
Ang III
2–8
Ang IV
3–8
A-(3–7)
Divalinal-
Ang IV
A-(1–5) ACE AP-A
AP-N
CP-P
PO
46
Centrally Active RAS Drugs
A large array of ACE inhibitors and ARBs have been approved for clinical use. These
drugs have differing levels of central activity depending on their ability to penetrate the
BBB.
Centrally active ACE inhibitors include: perindopril, ramipril, trandolapril, captopril,
fosinopril, lisinopril, prinivil, and monopril (Gao et al., 2013). However, some disagree as
to whether lisinopril and ramipril are centrally active (Li et al., 2010; O'Caoimh et al.,
2014).
Centrally active ARBs include: losartan, irbesartan, candesartan and telmisartan
(Michel et al., 2013). Candesartan and telmisartan also function as partial agonists of
peroxisome proliferator activated receptor gamma (PPAR-γ) (Erbe et al., 2006). Studies
have shown that the ARBs losartan, irbesartan, and candesartan attenuate the central
effects of Ang II on AT1 receptors, evident by an increase in mean arterial pressure,
drinking, and release of vasopressin (Culman et al., 2002). Of the three ARBs,
candesartan produced a 24-hour blockade at doses 5–10 times lower than losartan or
irbesartan (Gohlke et al., 2002). Furthermore, peripheral chronic administration of
candesartan (0.1, 0.5, or 1.0 mg/kg/day) demonstrated decreases in central Ang II
binding to AT1 receptors in a dose-dependent manner (Nishimura et al., 2000).
Candesartan delivered by p.o. route blocked the hypertensive effects of Ang II infusion
in rats. Candesartan also appeared to downregulate the brain RAS, as evident by the
decrease in angiotensinogen and ACE mRNA levels in the brain (Pelisch et al., 2010).
This suggests that candesartan can modulate the brain RAS. Therefore, candesartan
47
serves as a potent candidate to inhibit pathological AT1 signaling in the brain due to its
effective penetration of the BBB.
There were early concerns that ACE inhibitors would inhibit the ability of ACE to act
upon Aβ as a substrate, leading to diminished Aβ-degrading enzymatic activity, greater
Aβ load, and faster cognitive decline. However, clinical studies have determined these
concerns to be unfounded. Administration of the ACE inhibitor captopril in two lines of
APP transgenic mice did not show any changes in cerebral Aβ levels, plaque
deposition, or peripheral Aβ levels (Hemming et al., 2007).
Most studies have found either no effect or a beneficial effect of ACE inhibitors in
cognitive decline and dementia. Study results with ACE inhibitors have been mixed,
depending on the study design, patient population (e.g. older adults vs. demented
elderly), outcomes (e.g. incident dementia, rate of cognitive decline), and types of ACE
inhibitors. One study suggested that ACE inhibitors overall as a class were not
associated with incident dementia risk or cognitive decline in older hypertensive adults
(Sink et al., 2009). An observational case-control study demonstrated that dementia
patients showed a slower rate of cognitive decline after 6 months of treatment on a
centrally active ACE inhibitor (Gao et al., 2013).
Brain RAS Summary
The central RAS remains a topic of evolving ideas, but also a topic of many
uncertainties and undiscovered paths. The central RAS certainly mediates
cardiovascular pressor effects mainly through sympathetic output, vasopressin release,
and baroreflex control. Angiotensinergic pathways have also been described in their role
48
as neurotransmitters. Recent evidence suggests that Ang-(1–7)/Mas and Ang IV/AT4
play fundamental roles in learning and memory consolidation, and may be implicated in
cognitive decline and dementia.
However, the central RAS is clouded with uncertainties as well. The abundance of
angiotensinogen, Ang II, and processing enzymes in the brain point towards obvious
pathways that generate angiotensin metabolites. Low renin activity implies an alternate
catalytic pathway that forms Ang II from angiotensinogen. Recent findings suggest that
this renin-related activity may be due to the interaction between prorenin and its
receptor.
Central RAS enzymes and receptors appear to localize to both discrete
neuroanatomical regions and neuroglial cells. It is not entirely clear what the role of
neuroglial RAS elements entail yet beyond cardiovascular control, and whether there is
an emergent system.
The central brain RAS appears to be highly conserved across most mammalian
species for some elements of the RAS, such as receptor distribution. However, some
elements of the RAS differ by species, such as ACE expression in the choroid plexus,
which differs in rats versus humans.

49
Receptor Ligands Major responses
AT1 Ang II, Ang III Vasoconstriction,
increased sympathetic
output, vasopressin
release, decreased
baroreflex response
AT2 Ang II Vasodilation
Mas Ang-(1–7) Vasodilation, anti-
oxidative stress,
cerebroprotection, LTP
facilitation, learning and
memory
AT4 Ang IV, Ang-(3–7) Learning and memory

Table 2 Central angiotensin receptors, ligands, and major responses in the CNS.
Adapted from (Huijing Xia, 2008; Wright and Harding, 2013; Wright et al., 2013).

RAS Dysregulation in AD
The RAS has been shown to be dysregulated in AD. Clinical evidence suggests that
antihypertensive therapies that modify the production or actions of Ang II may protect
against AD. These may be related to blocking the pathological actions of Ang II, as Ang
II has been shown to exacerbate Aβ deposition and neurodegeneration. Alternatively,
pharmacological RAS intervention may be activating the central protective ACE2/Ang-
(1–7)/Mas pathway to elicit neuroprotective effects. Here I review the clinical and
preclinical studies of ARBs, Ang II, and Ang-(1–7) pertinent to cognitive decline and AD.

50
Clinical Studies of ARBs in AD
Midlife hypertension correlates with an increased risk of AD later in life (Launer et al.,
2000; Posner et al., 2002; Skoog et al., 1996; Yoshitake et al., 1995). Antihypertensive
therapies that modify the production or actions of Ang II have been shown to reduce
neurodegeneration and amyloid deposition. In a prospective cohort study of 819,491
predominantly male veterans with cardiovascular disease, ARBs were associated with a
significant reduction in the incidence and progression of AD and dementia as compared
to lisinopril or other classes of antihypertensives (Li et al., 2010). While ACE inhibitors
such as lisinopril modify the production of Ang II, ARBs modify the actions of Ang II. By
blocking Ang II binding to the AT1 receptors, ARBs increase available Ang II to either:
(1) bind to AT2 receptors or (2) further process to Ang-(1–7) and bind to Mas receptors.
The activation of protective AT2 or Mas signaling may be responsible for the lowered
incidence of AD and dementia.
Further, ARBs were shown to reduce amyloid deposition in patients at autopsy
(Hajjar, 2012). In this study of 890 hypertensive older adults, patients treated with ARBs
showed fewer amyloid deposition markers compared to those treated with other
antihypertensives. Patients with AD or other dementias also had fewer prescriptions for
ARBs and ACE inhibitors compared to other antihypertensives (Davies et al., 2011).
Patients with AD were half as likely to be prescribed ARBs.
Taken altogether, these data provide the strongest clinical evidence for ARBs
conferring additional protection against cognitive decline, suggesting another
mechanism beyond blood pressure control. The underlying mechanism may be related
51
to AT1 blockade, or further activation of alternate RAS signaling pathways such as AT2
and Mas.
Clinical studies have shown promising results for brain-penetrant ARBs in reducing
dementia, with candesartan showing the strongest dose-dependent reduction (Li et al.,
2010). However, there remains a multifaceted disconnect between preclinical and
clinical studies. This stems from the wide array of ARBs, their varying degrees of brain
penetration and dosing, and multiple AD models (Corbett et al., 2012).
Preclinical Studies of ARBs in AD
Preclinical studies using animal models have had mixed and discrepant results. This
may be due to the length of treatment, delivery route, and choice of ARB. Some ARBs
have pharmacological activity beyond AT1 blockade. For example, candesartan and
telmisartan can both penetrate the BBB and exert partial PPAR-γ agonism.
Initial in vitro screening studies identified valsartan as a viable drug candidate (Wang
et al., 2007). Valsartan was shown to reduce Aβ accumulation and inhibit Aβ
aggregation in vitro. Wang et al. demonstrated that preventive treatment with valsartan
for 5 months beginning at 6 months of age in Tg2576 mice reduced plaque burden and
improved performance on the Morris Water Maze task.
Ferrington et al. looked at whether ARBs, specifically valsartan and eprosartan, had
any effects on AD neuropathology and cognition in a 3xTg-AD mouse model (Ferrington
et al., 2012). They investigated two paradigms: early treatment (3–4 months of age) for
2 months, and treatment beginning at middle age (9–10 months of age) for 6 months.
Drugs were delivered by drinking water. No effects on APP, Aβ, tau, or performance on
52
T-maze or Morris Water Maze were observed in either paradigms. These negative
results may have been due to the choice of drug delivery route by drinking water.
Several factors may have played in role, including but not limited to: drug stability (which
was not evaluated), poor choice of ARBs, gastrointestinal degradation, and inconsistent
dosing.
Another study investigated whether candesartan can protect against scopolamine
(SCO)-induced memory impairment in rats, a model based on the cholinergic
hypothesis (Tota et al., 2012). Rats were treated for a week with candesartan (p.o.)
and/or PD123,319 (s.c.) prior to SCO administration. SCO administration resulted in
poor performance on the Morris Water Maze task, reduced CBF and ACh, along with
elevated AChE activity. Candesartan prevented SCO-induced amnesia and restored
CBF and ACh levels, which was blunted by coadministration with PD123,319.
PD123,319 by itself did not exacerbate the negative effects of SCO treatment. These
results suggest that candesartan’s neuroprotective effects may be partially mediated
through AT2 signaling and the cholinergic system.
An intriguing study found that intranasal losartan at sub-antihypertensive doses (10
mg/kg every other day) reduced Aβ plaques by 3.7-fold in 7-month-old APP/PS1
transgenic mice after 2 months of treatment (Danielyan et al., 2010). Losartan-treated
APP/PS1 mice exhibited diminished GFAP immunoreactivity in the entorhinal cortex and
hippocampus compared to vehicle-treated APP/PS1 mice. Aβ reduction was
accompanied by decreases in serum levels of GM-CSF, IL-1β, and IL-12p40/p70, and
an increase in IL-10, an anti-inflammatory cytokine. No significant changes were noted
53
in tail cuff blood pressure after 10 days of treatment. These results suggest that the Aβ-
lowering properties of intranasal losartan may be related to anti-inflammatory effects
independent of blood pressure control.
Overall, preclinical studies suggest that ARBs may prevent or delay AD
neuropathology. These depend on administering centrally active ARBs, or delivering
ARBs to bypass the BBB such as intranasal administration. While the specific
underlying mechanisms are unclear, they may be mediated through anti-inflammatory
effects and potentially even the cholinergic system.
Ang II in AD
ARBs have been shown to improve cognitive outcomes and stave Aβ deposition in
both preclinical and clinical studies, providing indirect evidence that pathological Ang II-
AT1 signaling may be the main contributing factor. Whether the underlying mechanism
is through blockade of Ang II’s action through AT1 activation is still unclear.
Studies investigating whether Ang II directly modifies Aβ deposition have been
mixed. Initial in vitro studies found that Ang II had no effect on secretase activity or Aβ
production in either primary hippocampal neurons and or human embryonic kidney 293
(HEK293) cells transfected with human AT1 and APP/PS1 (Wang et al., 2011).
However, intracerebroventricular infusion of Ang II in rats showed modulated expression
levels of APP and secretases, and increased Aβ40 and Aβ42 production (Zhu et al.,
2011). Increasing doses of Ang II was accompanied by concomitant increases in mRNA
levels of BACE1 and PSEN1. These effects were blocked by coadministration with
losartan, which suggests that these effects were mediated through AT1 activation.
54
Importantly, these results show that ARBs can modify and inhibit AT1-mediated Aβ
production. The authors suggest that increased central Ang II, essential to stress
response, may exacerbate amyloidogenic processing and Aβ deposition. However, it
should be noted that this study did not use a transgenic model in which Aβ readily
aggregates into plaques.
The deleterious effects of Ang II in the brain have been well described; it can induce
cerebrovascular remodeling, promote vascular inflammation and oxidative stress, and
impair the regulation of cerebral blood flow (Mogi et al., 2012). AT1 activation impairs
neurovascular coupling and drives vascular senescence (Kazama et al., 2004). The
neurovascular hypothesis proposes that Aβ clearance across the BBB may be impaired
due to aberrant angiogenesis or senescence of endothelial cells, resulting in diminished
efflux of Aβ through transporters such as LRP1 (Zlokovic, 2005). As a result, chronic
AT1 activation can drive endothelial dysfunction and neurovascular uncoupling, and
impair the ability of endothelium-resident efflux transporters to clear Aβ across the BBB.
Animal studies have shown that candesartan can protect against SCO-induced
memory impairment that coincide with reduced CBF and ACh levels, which suggests
Ang II may interfere with learning and memory (Tota et al., 2012). Peripheral and central
administration of Ang II via subcutaneous and intracerebroventricular routes significantly
impaired the ability of rodents to perform on spatial memory tasks such as the Morris
Water Maze and Elevated Plus Maze (Duchemin et al., 2013; Tota et al., 2013).
Peripheral administration of Ang II also coincided with reduced CBF and ACh levels.
Coadministration with candesartan prevented Ang II-induced memory impairment and
55
significantly ameliorated reductions in CBF and ACh levels. Earlier studies showed that
Ang II blocked hippocampal LTP, which represents a potential mechanism underlying
the learning and memory deficits (Denny et al., 1991). Taken together, these studies
suggest that Ang II interferes with learning and memory through AT1 activation, an effect
that may be partially mediated through interactions with the cholinergic system.
Ang-(1–7) in AD
ARBs directly modulate the pathological actions of Ang II by blocking the AT1
receptor. Alternatively, blockade of the AT1 receptor can lead to more available Ang II to
be cleaved by ACE2 to Ang-(1–7), which can then bind to the Mas receptor. The
protective ACE2/Ang-(1–7)/Mas pathway has been implicated in its role in brain
inflammation and a range of neurodegenerative diseases, including AD (Mogi and
Horiuchi, 2009; Saavedra, 2012; Wright et al., 2013; Xu et al., 2011). Studies suggest
that Ang-(1–7) may play a role in learning and memory, a finding supported by
preclinical and clinical studies showing increased cognitive benefit with activation of the
protective RAS axis.
Ang-(1–7) in Learning & Memory
While many studies have investigated the effects of ACE inhibitors and ARBs in AD,
few studies have looked at the direct effects of Ang-(1–7). Early studies in the 1990s
have shown that Ang-(1–7) may be implicated in learning and memory. Ang-(1–7)
enhanced LTP in the CA1 region of the hippocampus, while coadministration with the
Mas antagonist A-779 blocked this effect. Genetic deletion of the Mas receptor also
abolished this LTP-enhancing effect (Alenina et al., 2008; Hellner et al., 2005). LTP
56
deficits in the hippocampus form one of the cellular mechanisms that affect learning and
memory in AD. Soluble Aβ fragments disrupt LTP, leading to early cognitive decline
(Rowan et al., 2003). Since Ang-(1–7) affects LTP, an underlying cellular mechanism of
learning and memory, Ang-(1–7) may play a role in cognitive decline and aging.
Epidemiological studies have identified diabetes as a significant risk factor for AD
pathogenesis (SIMA, 2010). Animal models of diabetes exhibit cognitive deficit,
commonly measured by behavioral maze tasks. Studies have shown that Ang-(1–7)
treatment ameliorates cognitive deficits in rodent models of diabetes. In a
streptozotocin-induced rat model of type 1 diabetes, Ang-(1–7) has been shown to
improve their performance on the Morris Water Maze, increase GDNF and GFAP
expression, lower caspase-3 expression, and increase the number of surviving neurons
in the hippocampus (Zhang et al., 2015). These beneficial effects of Ang-(1–7) were
blocked by coadministration with the Mas antagonist A-779.
Preclinical Studies of Ang-(1–7) in AD
Recent studies have begun to investigate the role of Ang-(1–7) in mouse models of
AD. Ang-(1–7) levels in the brain have been shown to be reduced in the disease
progression of senescence-associated mouse prone 8 (SAMP8) mice, a model of
sporadic AD. A significant inverse correlation was observed between brain Ang-(1–7)
levels and tau hyperphosphorylation. This inverse relationship was also confirmed in
P301S mice, a model of pure tauopathy (Jiang et al., 2015). These findings suggest that
Ang-(1–7) may be implicated in the etiology and progression of AD, possibly by
modulating tau hyperphosphorylation.
57
Another group studied the therapeutic effects of intracerebroventricular (ICV)
infusion of Ang-(1–7) in 5xFAD mice (Uekawa et al., 2016). The 5xFAD mouse is an
aggressive mouse model of AD that develops amyloid deposition and cognitive deficits
at as early as 2 months of age (Chin, 2011). Seventeen-month-old male 5xFAD mice
were implanted with an ICV infusion cannula and osmotic pump, receiving either: (1)
artificial CSF vehicle, (2) Ang-(1–7) at 500 ng/kg/h, or (3) Ang-(1–7) and A-779 at 5.0
µg/kg/h for 4 weeks. Ang-(1–7) treatment significantly ameliorated cognitive impairment
on the Morris Water Maze task. Cerebral blood flow reactivity was also enhanced by
Ang-(1–7), as measured by an acetazolamide-induced increase in cerebral blood flow.
These beneficial effects of Ang-(1–7) were abolished by coadministration with A-779.
Although the 5xFAD mice exhibited significant amyloid deposition at 17 months of age,
treatment showed no statistically significant difference in hippocampal or cerebral Aβ
deposition, soluble Aβ42, or soluble Aβ oligomers.
Cognitive impairment can also manifest in other disease states and conditions.
Patients with congestive heart failure (CHF), concomitantly diagnosed with cognitive
decline and memory loss, exhibit increased hospital readmission rates and mortality.
Cognitive impairment can be elicited in a CHF mouse model by ligation of a coronary
artery to induce myocardial infarction (MI). After 8-weeks post-MI, CHF mice displayed
significant impairment on the spatial memory and recognition memory tasks, Morris
Water Maze and NOR. Treatment with Ang-(1–7) (50 µg/kg/hr) subcutaneously for 3
weeks significantly improved NOR discrimination ratios and spatial memory (Hay et al.,
58
2017). These results suggest that systemic administration of Ang-(1–7) may be a useful
therapeutic in treating symptoms of cognitive impairment in CHF.
Clinical Studies of Ang-(1–7) in AD
Emerging clinical evidence is beginning to show the involvement of the protective
ACE2/Ang-(1–7)/Mas axis in AD. ACE2 is a zinc metallopeptidase that generates Ang-
(1–7) from Ang II, and to lesser degree Ang-(1–9) from Ang I. A 2014 study showed that
serum ACE2 activity was reduced in AD patients compared to control subjects (Liu et
al., 2014). Further, this study also found that ACE2 could cleave Aβ43, an early
deposited form of Aβ that is highly amyloidogenic, to Aβ42. Aβ42 could then be broken
down further to non-toxic species such as No change was observed in serum Ang II
levels. These findings spurred further investigation into ACE2 activity within the brain
and its relation to AD pathology.
Post-mortem brain samples from AD patients demonstrated significantly lower ACE2
activity (approximately 50%) compared to age-matched controls, which inversely
correlated to Aβ levels and p-tau pathology (Kehoe et al., 2016). When AD patients
were stratified into Braak tangle stages, ACE2 was found to be significantly lower in the
advanced stages V–VI compared to stages 0–II. ACE2 was also significantly lower in
patients with an APOE ε4 allele. Brain Ang II levels were found to be significantly
elevated in AD patients, however Ang-(1–7) levels were unchanged. As a proxy
measure of ACE2 activity, the Ang II/Ang-(1–7) levels were significantly increased in AD.
These data suggest the conversion of Ang II to Ang-(1–7) is lower in AD likely due to
reduced ACE2 activity.
59
Based on the prior work that showed that brain Ang-(1–7) inversely correlated with
tau pathology in mice (Jiang et al., 2015), Jiang et al. investigated plasma Ang-(1–7)
and ACE2 activity in AD patients (Jiang et al., 2016). They found that plasma Ang-(1–7)
levels were significantly reduced in AD patients compared to their age- and gender-
matched controls (15.53 ± 4.35 pg/mL vs. 19.58 ± 3.22 pg/mL, p < 0.001) by ELISA
(Jiang et al., 2016). Plasma Ang-(1–7) positively correlated with MMSE scores. Their
receiver-operating characteristic (ROC) analysis revealed that plasma Ang-(1–7) levels
could distinguish AD patients from control subjects with a sensitivity and specificity of
69.1% and 74.2%, respectively, using an optimal cutoff of 18.2 pg/mL. Further, AD
patients showed a trend in reduced plasma ACE2 activity (p = 0.1729). This was the first
evidence that plasma Ang-(1–7) could be a potential biomarker for AD diagnosis.
In summary, the ACE2/Ang-(1–7)/Mas pathway is beginning to show some
interesting, albeit discrepant findings, in the context of AD. ACE2 activity appears to be
dysregulated in AD patients, more so in the brain than in circulation. Brain ACE2 activity
was significantly reduced in post-mortem AD brains, and was inversely correlated with
Aβ and p-tau burden (Kehoe et al., 2016). However, brain Ang-(1–7) levels remained
unchanged. Two studies showed a tendency, although statistically insignificant, towards
decreased circulating ACE2 activity in the serum and plasma of AD patients (Jiang et
al., 2016; Liu et al., 2014). Simultaneously, plasma Ang-(1–7) levels were found to be
significantly decreased in AD patients (Jiang et al., 2016).
Based on the recent evidence, there is a dysregulated local RAS within the AD brain,
which is reflected in part by the traditional circulating RAS. ACE2 catabolizes multiple
60
substrates within the brain, including Ang II and Aβ43. The finding that ACE2 can break
down Aβ43 is a significant one, as it adds another facet of the RAS within the context of
the Aβ cascade hypothesis. It is possible that diminished brain ACE2 activity is the
result of non-RAS related substrate activity (i.e. Aβ43). This was reflected in the
heightened levels of brain Ang II in AD patients, while Ang-(1–7) levels remain
unchanged.
RAS and PPAR-γ Activity
Beyond AT1 blockade, ARBs may elicit additional pharmacological actions, including
partial PPAR-γ activity. This partial PPAR-γ agonism of select ARBs may yield additional
neuroprotection (Saavedra, 2012). PPAR-γ is an intracellular nuclear hormone receptor
that regulates the expression of pro-inflammatory genes including inhibition of pro-
inflammatory transcription factors AP-1 and NF-κB. PPAR-γ have been well studied for
their peripheral actions and insulin sensitization in type 2 diabetes. Due to their anti-
inflammatory actions in the CNS, PPAR-γ are currently being studied for their utility in
neurodegenerative diseases with an inflammatory component such as AD (Heneka et
al., 2011).
PPAR-γ activation has been shown to modulate the activation state of macrophages,
shifting them from a pro-inflammatory M1 phenotype to an anti-inflammatory, phagocytic
M2 phenotype (Bouhlel et al., 2007). Traditional PPAR-γ agonists such as
thiazolidinediones have been shown to significantly reduce Aβ levels in transgenic AD
animal models, possibly associated with polarization of microglia towards M2 phenotype
(Mandrekar-Colucci et al., 2012). This finding was associated to an increase in Liver X
61
Receptors (LXRs) to increase LXR target genes, ABCA1 and ApoE. Lipidation of ApoE
by ABCA1 enhances the Aβ-binding properties of ApoE (Wildsmith et al., 2013).
Consequently, lipidated ApoE binds Aβ to prevent aggregation and facilitate proteolytic
degradation of Aβ by IDE and NEP (Jiang et al., 2008). Candesartan and Ang-(1–7)
both activate the PPAR-γ pathway.
This increase in ABCA1 and ApoE levels may assist glial phagocytosis by mediating
Aβ transport into microglia and astrocytes. Reduction in Aβ plaques were also
associated with the appearance of Aβ-laden microglia and astrocytes, suggesting that
PPAR-γ activation directly stimulates Aβ glial phagocytosis (Mandrekar-Colucci et al.,
2012).
Of note, PPAR-γ activity is not necessarily independent of AT1 blockade (Saavedra,
2012). Crosstalk properties between PPAR-γ activation and AT1 receptors exist. PPAR-
γ agonists reduce AT1-mediated inflammation, and downregulate AT1 receptor
expression. Conversely, AT1 activation with Ang II can also downregulate PPAR-γ
activity. Thus, AT1 and PPAR-γ signaling pathways may inhibit each other. Additionally,
Ang-(1–7) may also stimulate PPAR-γ activation, possibly through the induction of
endogenous ligands that bind to PPAR-γ or through upregulation of PPAR-γ itself
(Dhaunsi et al., 2010; Mario et al., 2012).
ARBs with partial PPAR-γ activity such as candesartan or telmisartan may be
beneficial in AD due to their anti-inflammatory actions and ability to modulate glial
phagocytosis. Traditional PPAR-γ agonists such as thiazolidinediones do not penetrate
62
the BBB effectively, and are associated with adverse effects such as edema, weight
gain, and cardiovascular risk (Rizos et al., 2009).

Endothelial Progenitor Cells
Endothelial progenitor cells (EPCs) were first described in 1997 as a progenitor cell
in the endothelial lineage derived from bone marrow (BM) (Asahara et al., 1997). EPCs
participate in post-natal vasculogenesis and migrate to ischemic sites of injury. Since
the initial discovery, EPCs have been implicated in a number of pathological conditions.
In patients at risk for cardiovascular disease, the number of circulating EPCs negatively
correlated with the Framingham cardiovascular risk score. EPCs isolated from patients
with diabetes mellitus type 2 exhibited decreased proliferative and migratory capacity in
vitro. More recently, reduced numbers of functional EPCs have been observed in
patients with AD. Enumeration and functional characterization of EPCs have spurred
further investigation into its utility as a biomarker or therapeutic intervention for disease.
Methods of Isolation, Enumeration, and Proliferation
Putative human EPCs have been identified, isolated, and characterized by two
techniques: flow cytometry and colony-forming assays (Basile and Yoder, 2014; Luong
and Gerecht, 2009; Yöder, 2012). Modern polychromatic flow cytometry allows for high
throughput analysis of single cells, suitable for measuring rare cell populations like
EPCs. Briefly, peripheral blood mononuclear cells (PBMCs) are isolated from
anticoagulated whole blood by density gradient separation, stained with fluorochrome-
labeled antibodies against EPC markers, and analyzed on a flow cytometer. Three
63
commonly used cell surface antigens for early EPCs include CD34, CD133, and
VEGFR-2. CD34 and CD133 are hematopoietic progenitor cell markers, while VEGFR-2
(also known as vascular endothelial growth factor receptor 2, CD309, KDR [human],
Flk-1 [rodent]) is an endothelial cell marker. CD133 typically co-expresses with
CD34
bright
, suggesting it is a marker for an early progenitor. As EPCs mature into
endothelial cells, they lose CD133 expression (Hristov et al., 2003). Using flow
cytometry, it may be possible to not only quantify EPCs, but also assess differing stages
of maturation based on their antigen expression.
EPCs have been cultured and expanded in vitro via differing techniques, yielding cell
populations with differing phenotypes. Thus, these cultured “EPCs” are better classified
by their respective isolation methods. These isolation systems have yielded at least
three populations: (1) circulating angiogenic cell, (2) colony forming unit-Hill, and (3)
endothelial colony forming cell.
Initially described in the seminal EPC paper, circulating angiogenic cells (CACs)
involve plating PBMCs in endothelial growth medium on fibronectin-coated plates. After
5–7 days of culture, adherent cells that ingest acetylated low-density lipoprotein
(AcLDL) and bind to lectin are designated as CACs. However, platelet microparticles
frequently contaminate PBMC preparations, which may lead to false positives since
platelets share many cell surface markers with endothelial cells. Further, adherent
monocytes cultured with vascular endothelial growth factor (VEGF) share many features
with EPCs, including AcLDL uptake and expression of VEGFR-2, CD31, and von
Willebrand factor (vWF). CACs are likely a heterogeneous population composed of
64
EPCs and angiogenic monocytes of hematopoietic lineage, which all support
vasculogenesis. Putative EPC colonies consist of a central core of rounded cells
surrounded by radiating spindle-shaped cells.
The colony forming unit-Hill (CFU-Hill) assay involves preplating PBMCs on
fibronectin-coated plates for 2 days, then replating the non-adherent cells. This replating
step attempts to mitigate the shortcomings of the first CAC isolation system by removing
adherent monocytes, macrophages, and mature endothelial cells. At day 5, colonies are
scored as CFU-Hill. This assay can now be performed using a commercially available
kit, the CFU-Hill Liquid Medium Kit from Stem Cell Technologies. In patients, CFU-Hill
colonies were reported to be inversely correlated with the Framingham cardiovascular
risk factor score (Hill et al., 2003).
A third clonogenic assay measures endothelial colony forming cells (ECFCs) based
on their proliferative potential. PBMCs are cultured on plates coated with type 1
collagen. After 2–3 weeks, adherent colonies become visible. These colonies can then
be replated repeatedly for clonal expansion or characterization. Colonies obtained in the
ECFC assay are significantly rarer than the previous methods, estimated at ~1 ECFC
per 20 mL of whole blood (Mead et al., 2008).
In summary, these EPC culture methods yield cells that clearly show angiogenic
potential, but differ in cell/colony number, length of culture, and proliferative potential.
Both CAC and CFU-Hill assays display limited replating and proliferative capacity
compared to ECFC isolations, but require less time to culture.
65
There is controversy as to which isolation method yields true EPCs. Research has
shown that cells from CAC and CFU-Hill preparations may in part be derived from
hematopoietic monocytes and macrophages, which support vasculogenesis but do not
directly differentiate into endothelial cells (Yoder et al., 2007). Therefore, some argue
that CAC and CFU-Hill preparations do not yield true endothelial progenitors. Despite
this, putative EPCs defined from flow cytometry or CAC/CFU-Hill assays may still serve
as useful biomarkers to examine cardiovascular risk or endothelial dysfunction.
EPCs in AD
The role of EPCs in AD has only begun to be investigated in the past decade. The
first report by Lee et al. in 2009 showed that patients with AD show reduced numbers of
CACs, after matching for risk factors including age, sex, and Framingham risk score
(Lee et al., 2009). Patients with AD had fewer CACs, lower Mini-Mental State
Examination (MMSE) scores, and higher Clinical Dementia Rating (CDR) compared to
risk-factor controls. However, there were no differences in circulating CD34+VEGFR-2+
or CD133+ cell counts measured by flow cytometry, and these counts did not correlate
with CACs.
In a follow-up 2010 study, Lee et al. examined the functional characteristics of CACs
in AD (Lee et al., 2010). They found that CACs from AD patients showed reduced
chemotaxis, increased senescence-associated β-galactosidase activity, and reduced
angiogenic ability on the Matrigel assay.
Another report by Kong et al. in 2011 also looked at circulating EPCs by flow
cytometry. Based on the finding by Lee et al. that observed no changes in
66
CD34+VEGFR-2+ cells or CD133+ cells, instead Kong et al. focused on measuring
CD34+CD133+ cells as a marker for EPCs. They found that patients with AD had
significantly lower circulating CD34+CD133+ cells than control subjects, which
correlated with a decrease in MMSE scores. Patients with VaD also had lower
CD34+CD133+ cells, but this did not correlate with MMSE scores.
Although there is clinical evidence of diminished circulating EPCs in AD, little is
known of the circulating levels or impact of EPCs in animal models of AD. A recent 2014
study by Safar et al. investigated the treatment with BM-derived EPCs in a rat model of
SCO-induced cognitive impairment. Repeat dosing with SCO results in symptoms akin
to AD pathogenesis, including learning & memory deficits that coincide with the
accumulation of Aβ and p-tau. Safar et al. obtained BM-EPCs by culturing BM-derived
mononuclear cells from rats on fibronectin-coated plates for 7 days. A single intravenous
administration of BM-EPCs at day 5 into the 6-week chronic daily SCO treatment
resulted in positive staining for fluorescently labeled BM-EPCs in the rat hippocampus,
indicating successful EPC migration into the hippocampus. BM-EPC treatment showed
significant improvements on spatial memory tasks, including the Morris Water Maze and
spontaneous alternations in the Y-maze test, and attenuated levels of Aβ42 and p-tau.
The group also found that mRNA expression of APP and GSK-3β decreased, while NEP
mRNA expression increased. This suggests that BM-EPC treatment may be modifying
amyloid production and/or catabolism. Despite these positive findings, it is surprising
that circulating EPCs were not measured.

67
Effects of Modulating the RAS on EPCs
There has been increasing interest in modulating EPC levels through
pharmacological intervention, particularly through modulation of components in the
RAS. Patients with hypertension are frequently prescribed ACE inhibitors or ARBs,
which modify the production or action of Ang II on its constitutive receptor.
Several studies have shed light on the influence of Ang II on EPCs. Ang II has been
shown to upregulate VEGFR-2 expression in human EPCs in vitro, possibly potentiating
EPCs through VEGF stimulation (Imanishi et al., 2004). However, Ang II may also have
deleterious effects on the proliferative capacity of EPCs by driving them towards
senescence via free radical formation, marked by increased β-galactosidase staining
and diminished telomerase activity (Imanishi et al., 2005). These effects were reversed
by coadministration with valsartan in vitro. It has been hypothesized that acute AT1
stimulation can initially lead to pro-angiogenic effects and EPC recruitment, but that
chronic AT1 stimulation may lead to diminished EPC numbers over time (Durik et al.,
2012).
ARBs have been shown to increase circulating EPCs in patients with diabetes
mellitus type 2 (Bahlmann et al., 2005; Reinhard et al., 2010), coronary artery disease
(Endtmann et al., 2011; Pelliccia et al., 2010), acute coronary syndrome (Porto et al.,
2009), and kidney transplants (Townamchai et al., 2010). Animal studies have also
shown similar effects of ARBs in models of hypertension (Yao et al., 2007; Yoshida et
al., 2011; Yu et al., 2008) and hind limb ischemia (Wang et al., 2006a; You et al., 2008).
68
Along with the in vitro studies, these human and animal data present strong evidence
for using ARBs to bolster the pool of circulating EPCs.
Some of the beneficial effects seen in ARBs may also be partially due to the actions
of Ang-(1–7). While Ang II exerts its actions primarily through the AT1 receptor, blockade
of the receptor can lead to further processing to Ang-(1–7), which then binds to the
putative Mas receptor. Studies have shown that Ang-(1–7) improved the recovery of
hematopoietic stem cells, from which EPCs are derived (Ellefson et al., 2004; Heringer-
Walther et al., 2009; Rodgers et al., 2012; 2005; 2002; 2003). Therefore, it is plausible
that the actions of Ang-(1–7) may affect EPC numbers.
Studies have begun to investigate whether Ang-(1–7) directly stimulates EPCs. Ang-
(1–7) was shown to increase proliferation of cultured AcLDL+/lectin+/VEGF-R2+ BM-
EPCs from rodents in a dose-dependent manner after 7 days, while coadministration
with the Mas receptor antagonist A-779 blocked this effect (Wang et al., 2010b). In a
mouse model of type 2 diabetes mellitus, 14-day administration of Ang-(1–7)
significantly increased circulating Sca-1+/Flk-1+ EPCs measured by flow cytometry
(Papinska et al., 2015). Two mechanisms of action have been proposed: (1) activation
of eNOS through Akt signaling may be a possible pathway for VEGF-induced
angiogenesis, and (2) Ang-(1–7) may inhibit AT1-mediated NADPH oxidase radicals,
which drive EPC senescence (Roks et al., 2011). Whether circulating Ang-(1–7) levels
are associated with EPCs in humans has yet to be investigated. Further work must be
conducted to elucidate the role and mechanism of Ang-(1–7) in EPCs.

69
Progenitor Cells in AD
There is growing evidence that there are progenitor cell deficits in AD beyond normal
aging. These deficits may reflect poor reserves of hematopoietic stem or progenitor
cells in the BM, or blunted/dysregulated mobilization or differentiation to sites of injury.
Clinical trials have commenced using hematopoietic growth factors as a therapy for AD.
Preclinical studies suggest that these growth factors may induce myeloid infiltration into
the brain, and increase phagocytosis of pathogenic Aβ by both BM-derived
macrophages and brain-resident microglia. Here I review hematopoiesis, hematopoietic
growth factors, adult neurogenesis, and the current lines of evidence for observational
and interventional studies in progenitor cells in AD.
The Hematopoietic System
Hematopoiesis refers to the production of blood cells in the body. In embryonic
development, blood cells emerge from a multipotent hemangioblast cell. Primitive
hemangioblasts differentiate from other embryonic stem cells by the expression of ACE
(CD143) (Zambidis et al., 2008; 2007). Hemangioblasts gives rise to hematopoietic and
endothelial progenitors. These progenitor cells are thought to express CD34 and CD133
(Loges et al., 2004). Primitive hematopoietic structures emerge in the embryo called
blood islands. During fetal development, most hematopoiesis occurs in the liver and
then transitions to the BM. Adult hematopoiesis primarily occurs in the BM medulla.
BM can be found in nearly every large bone in the body. In children, hematopoiesis
primarily occurs in the femur and tibia. As humans age, hematopoiesis transitions to the
other bone structures, including: vertebrae, pelvis, sternum, and cranium.
70

Figure 7 Sites of Hematopoiesis in Human Development.
Illustration by Michał Komorniczak (Poland). Reprinted under Creative Commons
Attribution 3.0 Unported license.
Blood Cell Lineages
The hemangioblast cell diverges downstream to form endothelial progenitors and
hematopoietic progenitors. Hematopoietic progenitors give rise to erythrocytes,
leukocytes, and platelets. Progenitor differentiation is spurred and mediated by
hematopoietic cytokines. These are sometimes referred to as hematopoietic growth
factors or more accurately, hematopoietic colony-stimulating factors.
These colony-stimulating factors, either alone or synergistically with others, can
stimulate ex vivo colony growth. At present, it is difficult to ascertain the kinetics of
hematopoietic stem or progenitor cell (HSC/HPC) populations, therefore this information
must be inferred from clonal studies. Progenitors form colonies of maturing cells of
different lineages on a semisolid matrix (e.g. methylcellulose or agar) in the presence of
71
growth factors. As a result, these colonies and growth factors have been referred to as
colony-forming units (CFUs) and lineage colony-stimulating factors (e.g. granulocyte
colony-stimulating factor), respectively.
Identifying Progenitors by Cell Surface Antigens
Other than in vitro assays, progenitors have been isolated or defined by the
expression of certain cell surface antigens. The most commonly used cell surface
marker for hematopoietic stem and progenitor cells is CD34. CD34 cells comprise 2–5%
of nucleated cells in the BM, while they comprise 0.03–0.09% of leukocytes in
circulation. However, both HSCs and HPCs (as defined by functional colony-forming
assays) are rarer; HSCs occur in approximately 1 in 20,000 BM cells. CD34 is useful as
a marker to enrich for primitive cells, but are not specific only for HSCs and HPCs.
CD133 was first discovered to be co-expressed with a population of CD34
bright
cells,
which suggests that positive selection with early lineage markers (e.g. CD133) and
negative selection with mature markers (e.g. Lin, CD45) may help further enrich for
progenitors. Similar strategies have been utilized to define putative EPC populations by
polychromatic flow cytometry (Estes et al., 2010; Hristov et al., 2009; Mund et al., 2012;
Rustemeyer et al., 2006; Shaffer et al., 2006).
Progenitor Sources and Uses
There is no clear defined way to isolate a self-renewing multipotent stem cell from
the hematopoietic system. Rather, the presence of stem or progenitor cells have been
inferred by processes that enrich, or are known to be enriched, for stem cells. Enriched
72
stem cell populations exist in: harvested BM, mobilized peripheral blood (induced by G-
CSF or other stimulating factors), or umbilical cord blood.
These enriched stem and progenitor cells have major applications with tremendous
biomedical utility, including: allogeneic transplantation to restore hematopoiesis in BM
failure or genetic disease (e.g. aplastic anemia), autologous/allogenic transplantation to
reconstitute hematopoietic after radiochemotherapy, and in gene therapy to restore
defective stem cells (e.g. severe combined immunodeficiency [SCID])
Beyond the traditional hematopoietic system, other sources of stem cells have been
investigated, including: mesenchymal stem cells, human embryonic stem cells (hESCs),
and induced pluripotent stem (iPS) cells (Takahashi et al., 2007). The efficiency of the
therapeutic stem cells can be evaluated by the in vivo SCID repopulating cell (SRC)
assay in immunodeficient mice. It provides a surrogate, functional measure of long-term
hematopoietic reconstitution. Despite all the advances in our understanding of stem
cells, even hESCs and iPS cells show poor levels of engraftment on this assay. Further
advancement in this burgeoning field of therapeutic stem cells are still yet to come.
Role of RAS in Hematopoiesis
A plethora of evidence points to a deeply intertwined role of the RAS in the
development and regulation of hematopoiesis (Durik et al., 2012; Park and Zambidis,
2009; Roks et al., 2011). The role of the RAS takes root early in embryonic
development, in which hemangioblasts give rise to hematopoietic and endothelial
progenitors. Primitive hemangioblasts differentiate from other embryonic stem cells by
the expression of ACE (CD143) (Zambidis et al., 2007; 2008). Human embryoid body
73
(hEB)-derived ACE+ hemangioblasts exhibit significant AT2 upregulation, supporting the
notion that AT2 has a critical role in fetal development (Park and Zambidis, 2009;
Zambidis et al., 2008). Directing Ang II signaling with AT1 or AT2 inhibitors skew hEBs
differentiation towards endothelial cells or multipotent hematopoietic progenitors. These
findings suggest a balance of AT1 or AT2 signaling may drive progenitor cells to
differentiate into endothelial cells or proliferate further, respectively. Therefore, it may be
possible to direct the fate of hemangioblasts toward hematopoietic or endothelial
lineages by manipulation of the RAS.
Hematopoietic Growth Factors
Hematopoietic growth factors (HGFs) play a major role in the proliferation,
differentiation, and survival of primitive hematopoietic stem and progenitor cells. HGFs
belong to a family of glycoproteins, whose effects are mediated by high-affinity binding
of these growth factors to their cell surface receptors on target cells (Sieff, 2016).
Recombinant human HGFs have been developed for a variety of indications, mainly for
diseases or chemotherapy involving hematopoietic deficit such as myelodysplastic
syndrome and chemotherapy-induced neutropenia. Recombinant human HGFs include:
interleukin (IL)-3, IL-11, granulocyte-macrophage colony-stimulating factor (GM-CSF),
granulocyte colony-stimulating factor (G-CSF), erythropoietin (EPO), and
thrombopoietin (TPO).
In the clinic, G-CSF and GM-CSF have been used to stimulate hematopoiesis for
BM syndromes or peripheral blood mobilization. G-CSF and GM-CSF stimulate
granulocyte and granulocyte-monocyte lineages, respectively. EPO has been used to
74
stimulate RBC production for anemia. Other HGFs have been explored (e.g. SCF, IL-3,
TPO), but have found limited use due to safety or efficacy concerns. Further
development for these HGFs are shifting toward specific small-molecule agonists.
Interestingly, patients with polycythemia vera (i.e. extremely high levels of RBCs)
exhibit low serum EPO and may present with cognitive impairment, presumably
associated to a reversible decrease in cerebral blood flow due to elevated blood
viscosity. One patient case study showed that phlebotomy relieved cognitive
impairment, as determined by neuropsychological assessment (Di Pollina et al., 2000).
Altered levels of these growth factors exist in AD compared to healthy controls in
blood, cerebrospinal fluid, and brain, suggesting a dysregulated or compensatory
response (Sopova et al., 2014). Recombinant G-CSF (filgrastatim, Neupogen) and GM-
CSF (sargramostim, Leukine) are currently being investigated as a therapeutic for AD.
Both filgrastim and sargramostim are in Phase 2 for this indication. The mechanism of
action of G-CSF and GM-CSF in AD is proposed to increase phagocytosis of Aβ by
brain-resident microglia or BM-derived macrophages.
Non-hematopoietic Cells in the BM
Hematopoietic stem cells give rise to hematopoietic progenitor cells, which then
mature to effector blood cells, e.g. erythrocytes, leukocytes, and thrombocytes. Beyond
hematopoiesis, a small population of circulating BM-derived cells may take part in other
homeostatic functions including angiogenesis and vascular repair (Birbrair and Frenette,
2016). Under certain conditions, these progenitor cells can mobilize and differentiate to
other types of cells, such as endothelial cells, or more controversially neuronal and glial
75
cells (Jung et al., 2008; Shin et al., 2011). While these ideas may have been farfetched
in the past, the discovery of inducible pluripotent stem cells have challenged the
traditional notion of terminally differentiated cells (Takahashi et al., 2007). Cells can
undergo transdifferentiation into other cell types, in which mature somatic cells can
transform into another type of cell without undergoing an intermediate progenitor
phenotype (Badorff et al., 2003; Imamura et al., 2010; Ji et al., 2016).
Cognitive Decline in Disorders of Hematopoietic Deficit
Disease states with hematopoietic deficits have elicited clues into the role of these
progenitors in cognitive decline. Clinical studies in cases such as post-chemotherapy
have assessed neurological or cognitive outcomes as secondary measures (Janelsins
et al., 2011). These studies may yield insight into cognitive decline.
Hematopoietic deficits exist in conditions such as: chemotherapy-induced cytopenia,
radiation injury, and even amputation. Down syndrome (trisomy 21) is closely
intertwined with both early-onset dementia and hematopoietic deficit. Trisomy 21 is
characterized by the presence of a third copy of chromosome 21, which APP is present
on (Weksler et al., 2013). Patients with Down syndrome develop AD-like plaque
pathology by age 35 (McPhee and Hammer, 2009). Further, these patients exhibit
reduced numbers of lymphocytes and produce fewer antibodies, making them more
prone to infection (Hickey et al., 2012).
Other lines of evidence may lie in other neurological disorders, such as traumatic
brain injury (TBI). TBI is known to exacerbate the risk for AD, potentially through
incomplete resolution of inflammatory component (Breunig et al., 2013). Further
76
evidence for this may be elucidated in future studies to come, as the role of TBI in
military veterans and athletes is becoming more appreciated.
Adult Neurogenesis
It was long thought that the brain was a static organ in adulthood, with no cells
capable of self-renewal. A seminal paper by Eriksson, Gage, et al. unraveled the first
strings of evidence for adult neurogenesis in humans. Cancer patients received an i.v.
infusion of 5-bromo-2-deoxyuridine (BrdU), a fluorescent DNA intercalating agent and
thymidine analog, for diagnostic purposes (Eriksson et al., 1998). Postmortem brain
tissue revealed fluorescent cells in the hippocampus and subventricular zone,
suggesting that newly dividing cells incorporated BrdU. Since then, exciting lines of
research have flourished in the field of adult neurogenesis along with implications in
aging, cognitive decline, and dementia.
Neurogenic Niches
The environment and niche of adult neurogenesis is a complicated topic.
Particularly, where do these progenitor cells come from? What do they become? And do
they integrate into existing neural circuits?
Adult neurogenesis occurs in a few select areas in humans, including the
hippocampus and the subventricular zone. The hippocampus is an area of extreme
interest as it is the area most commonly affected first in amnestic cognitive impairment.
Adult neurogenesis can occur in the subgranular zone (SGZ) of the dentate gyrus (DG)
in the hippocampus. The SGZ is a layer of granular cells between the granule cell layer
and hilus of the dentate gyrus. Composed of neural stem cells, astrocytes, and
77
endothelial cells, the SGZ is a neurogenic niche that regulates and fosters proliferation,
migration, and differentiation.
The SGZ is an area of high vascularization. Endothelial cells provide a supporting
matrix for neurogenic cell attachment and paracrine factors that aid in proliferation and
growth, such as VEGF. Studies have also shown that angiogenic and neurogenic
pathways share common signaling pathways, which imply a reciprocal, synergistic, and
symbiotic relationship. These blood vessels can also transport hormones and other
signaling molecules from the peripheral circulation.
Evidence suggests that 700 new neurons are added daily in each hippocampus in
adult humans, corresponding to an annual turnover of 1.75% neurons in the renewing
fraction (Spalding et al., 2013). While there is a modest decline with normal aging, it
remains to be seen whether deficits in adult neurogenesis may contribute to cognitive
decline and dementia overall.
Progenitor Cell Homing to the Brain
A very interesting proof-of-concept study was conducted to see whether mobilized
BM-derived cells could home to the brain (Shin et al., 2011). Injection of G-CSF and
AMD3100 (CXCR4 antagonist) mobilized BM-HPCs into circulation in a mouse model of
AD. Simultaneously, these mice received an intracerebral injection of SDF-1α to home
BM-HPCs to the brain. This treatment yielded improvements in spatial memory, despite
no change in Aβ deposition (Shin et al., 2011). Treatment also increased the number of
proliferating BrdU+NeuN+ cells in the dentate gyrus, which suggests the combination of
BM mobilization with SDF-1 homing may also promote hippocampal neurogenesis.
78
As a therapeutic approach, there are significant regulatory and safety hurdles in
using these multiple modalities. While the HPC mobilization factors are approved for
certain clinical indications (filgrastim, a recombinant G-CSF; plerixafor, a CXCR4
antagonist), recombinant SDF-1α is not approved for use in humans. Intracerebral
delivery of a chemokine is itself a significant development hurdle. However, as a proof
of concept, it is certainly an intriguing one that shows that mobilization and homing of
progenitor cells to the brain is possible.
Clinical Studies of Progenitor Cells in AD
Several studies have shown that circulating progenitor cells may be dysregulated in
AD. The first evidence of this was described by Maler et al. Patients with early AD
exhibited decreased circulating CD34+ cells, which negatively correlated with CSF Aβ42
and Aβ42/40 ratio (Maler et al., 2006). CD34+ cells also negatively correlated
significantly with age within the AD group, but not the controls, perhaps reflecting
accelerated aging and an exhaustion of progenitor cells.
Kong et al. also showed a significant reduction in CD34+CD133+ cells in AD
patients, a finding previously attributed to EPCs (Kong et al., 2011). Diminished
CD34+CD133+ cells may instead reflect a deficit in circulating early progenitors. Higher
CD34+CD133+ cells correlated with higher MMSE scores, suggesting that these cells
may have a protective role in cognition.
An opposite pattern appeared to emerge in late-stage AD. Stellos et al.
demonstrated that circulating CD34+ and CD34+CD133+ cells significantly increased in
patients with moderate-to-severe AD but not mild AD (Stellos et al., 2010). This
79
mobilization of circulating progenitors may be a compensatory response to injury in the
brain. This was supported by the fact that CD34+CD133+ cells inversely correlated with
MMSE scores and SDF-1 levels, a potent chemokine known to mobilize progenitors.
The disparity between studies depicting lower progenitor cells in early AD and higher
progenitor cells in late-stage AD may reflect contrasting roles in brain injury. Higher
circulating progenitors may be protective in early AD, while mobilization of CD34+ cells
in late-stage AD may indicate a response to injury.
Cerebrovascular injury can mobilize progenitor cells with high proliferative potential.
Jung et al. isolated “neuronal outgrowth cells” from the peripheral blood of human stroke
patients (Jung et al., 2008). Isolated PBMCs were grown ex vivo on 2% gelatin-coated
plates in EGM-2 media for 8 weeks, which resembled a proliferative colony-forming
assay method similar to ECFCs. Cells were selected based on their endothelial or
neuronal outgrowth phenotype, designated by their morphology and gene expression
profiles. Neuronal outgrowth cells were transplanted in a rat model of focal cerebral
ischemia. Transplanted cells persisted for over 6 months in ischemic brains. Isolation
efficiency of neuronal outgrowth cells were higher from stroke patients compared to risk-
factor controls, likely due to the increased mobilization of BM progenitors in response to
vascular injury.
While protective progenitor cells mobilize in response to cerebrovascular injury, this
may deplete their reserves. It has been shown that high numbers of CD34+ cells
mobilize into circulation within hours and days after ischemic stroke (Machaliński et al.,
2006; Paczkowska et al., 2005). Taguchi et al. followed patients with a history of
80
cerebral infarcts for a year (Taguchi et al., 2008). After a year, patients with fewer
circulating CD34+ cells at follow-up showed significant worsening of neurologic function.
The exhaustion of protective progenitor cells after cerebrovascular injury may
exacerbate cognitive decline.
No studies have yet investigated the numbers of progenitor cells in the earliest
detectable clinical phase of cognitive decline, MCI. Also, studies have not examined
whether these early progenitor cell deficits correlate with structural MRI-based
measures of brain atrophy. It may be prudent to investigate these progenitor cells in the
absence of concomitant cardiovascular disease or brain injury, as they may deplete
progenitor cell reserves.

Gaps in Knowledge
AD is an insidious neurodegenerative disease that initially manifests as memory
loss. Cognitive function progressively worsens over time, and ultimately results in death.
No clinically approved treatments can cure or delay the progression of the disease.
Modifiable risk factors such as midlife hypertension can increase the risk of
developing AD later in life. Antihypertensive therapies that block the actions of Ang II
such as ARBs may confer additional protection against cognitive decline. The
mechanisms underlying this additional protection may lie beyond controlling blood
pressure, but rather through the modulation of neuroprotective elements of the RAS
such as Ang-(1–7). Exploring drugs that modulate protective RAS actions in an animal
model of AD is a logical significant next step.
81
In addition, AD patients have reduced numbers of functional circulating angiogenic
cells. This may reflect diminished progenitor reserves or hematopoietic deficit. These
cells may exert pro-angiogenic effects and provide trophic support to brain parenchymal
cells. Alternatively, these cells may directly contribute to adult neurogenesis. It is
unknown what role these cells may play in cognitive decline, and whether their numbers
and function are also reduced in earlier phases of cognitive decline such as MCI.
Further work must be done to elucidate the role of these angiogenic cells across the
spectrum of cognitive decline and whether they correlate to brain structure and clinical
function.
This thesis dissertation consists of two main parts: (1) a preclinical study
investigating the efficacy of an ARB and Ang-(1–7) in 3xTg-AD mice, (2) and a clinical
study investigating circulating angiogenic cells in older adults with MCI. Chapter 2
describes the preclinical animal study and vascular mechanisms. Chapter 3 delves into
possible Aβ clearance mechanisms. Chapter 4 explains the method development for the
clinical study. Chapter 5 describes the clinical study and current findings.
82
Chapter 2: Angiotensin-(1–7) and Candesartan Reduce Aβ
Introduction
Modifiable vascular risk factors such as midlife hypertension can increase the risk of
developing AD later in life. Controlling modifiable risk factors such as hypertension may
protect against cognitive decline later in life. Beyond blood pressure control, ARBs may
confer additional protection against cognitive decline.
ARBs have been shown to significantly reduce the incidence of AD and amyloid
deposition markers compared to other antihypertensives (Hajjar, 2012; Li et al., 2010).
Patients with AD were also less likely to be prescribed ARBs compared to other
antihypertensives (Davies et al., 2011). This clinical evidence strongly supports a
protective role for ARBs in AD, and the underlying mechanism may be related to AT1
blockade, or further catabolism of Ang II to Ang-(1–7) by ACE2. Newer evidence has
pointed towards harnessing the protective RAS (ACE2/Ang-(1–7)/Mas) in
neurodegenerative diseases such as AD (Mogi and Horiuchi, 2009; Saavedra, 2012;
Wright et al., 2013; Xu et al., 2011).
AT1 activation is known to impair neurovascular coupling and drive vascular
senescence (Kazama et al., 2004). Cerebrovascular dysfunction resulting from chronic
AT1 activation may impair the ability of endothelium-resident efflux transporters to clear
Aβ across the BBB. Further, senescent endothelial cells impair functional hyperemia
and cerebral blood flow (Girouard and Iadecola, 2006).
Circulating EPCs maintain vascular homeostasis by mobilizing to sites of injury and
releasing pro-angiogenic cytokines (Iadecola, 2013). These vascular repair cells have
83
been shown to be dysregulated and dysfunctional in AD (Kong et al., 2011; Lee et al.,
2009; 2010). Chronic Ang II stimulation has been shown to deplete EPCs (Durik et al.,
2012), thereby possibly exacerbating cerebrovascular dysfunction by compromising
vascular repair mechanisms.
Further work needs to be done to identify the underlying mechanisms responsible for
the protective effect of ARBs against cognitive decline. Animal studies involving ARBs in
AD models have found conflicting results, likely due to the selection of ARBs, dose, and
delivery route (Corbett et al., 2012; Danielyan et al., 2010; Ferrington et al., 2012; Tota
et al., 2012; Wang et al., 2007). Clinical studies have shown promising results for brain-
penetrant ARBs in reducing dementia, with candesartan showing the strongest dose-
dependent reduction (Li et al., 2010). We sought to see whether RAS therapy could
potentially prevent or delay the progression of AD pathology in a transgenic mouse
model of AD, and whether circulating EPCs may play a role in maintaining
neurovascular coupling.
In this chapter, the efficacy of candesartan and Ang-(1–7), alone and in combination,
was evaluated in the 3xTg-AD mouse model. The experimental design investigated the
effects of AT1 blockade with candesartan and Mas activation with Ang-(1–7) on
cognition and pathology in this animal model. I hypothesized that AT1 blockade and
Mas activation would reduce AD neuropathology, increase EPCs, and restore cerebral
blood flow. Prevailing clinical evidence indicates that candesartan may reduce the
incidence of AD. Further preclinical work is required to reveal the underlying
mechanisms by recapitulating these effects in relevant AD models.
84
Methods
Animals
The National Institutes of Health Principles of Laboratory Animal Care were followed,
and the Department of Animal Resources at the University of Southern California
approved this study. Five-month-old male and female 3xTg-AD mice and their non-
transgenic (NonTg) controls were obtained as a generous gift from Dr. Roberta Brinton’s
laboratory. Food and water were available ad libitum, and all mice were kept on a 12-
hour light, 12-hour dark cycle.
Chemicals and Reagents
Candesartan cilexetil powder was purchased from Sigma-Aldrich and prepared in
vehicle solution consisting of 0.57% DMSO and 2% polysorbate in 0.9% saline. Ang-(1–
7) was purchased from Bachem and prepared according to Good Manufacturing
Practices.
Study Design and Timeline
Male and female 3xTg-AD mice and NonTg controls (n = 5–7/genotype/group) were
randomized into 4 treatment groups: vehicle, candesartan, Ang-(1–7), or a combination
of candesartan and Ang-(1–7).
Treatments consisted of daily subcutaneous injections of vehicle, 1 mg/kg
candesartan, or 500 µg/kg Ang-(1–7) dissolved in 0.9% saline. Treatment began at 5
months of age, prior to significant Aβ plaque deposition, and lasted until 13 months of
age.
85
Behavioral tasks were initiated at 9 months of age, beginning with the least aversive
or stressful tests. At 9–10 months of age, spontaneous alternations on the T-maze was
evaluated to assess working memory. At 11 months of age, the novel object recognition
test was conducted to assess working memory, novelty, and motor activity. At 12 months
of age, prepulse inhibition was evaluated to assess sensorimotor gating.
Doppler ultrasound measurements of carotid blood flow were conducted a week
prior to sacrifice at 13 months of age. Circulating Flk1+Sca1+ EPCs and hippocampal
Aβ levels were characterized by flow cytometry and immunoassay, respectively, on
tissues collected at necropsy.

Figure 8 Animal Study Design and Timeline.

Behavioral Tasks
The hallmarks of AD describe a progressive decline in cognition and memory.
Phenotyping of novel exploratory behavior rely on the innate tendency of rodents to
seek novelty in favor of familiarity, which reflects working memory. These behavioral
tests translate readily from rodent models to humans.
Spontaneous Alternations on the T-maze
Mice were placed individually in the starting arm of the maze and allowed to choose
one of two goal arms in the “sample phase.” After the mouse entered one of the arms,
Age (months)
Treatment
daily subcutaneous injections for 8 months
9–10
T-maze
11
NOR
13
Carotid Blood Flow
Tissue Collection
EPC Flow Cytometry
12
PPI
0 5
Begin Treatment
86
the door to the corresponding chosen arm was closed. After a retention period of 30
sec, the mouse was replaced in the starting arm position and allowed to choose
between two open goal arms in the “choice phase.” Successful alternations were scored
when the mouse chose opposite arms in the sample and choice phases. Both phases
were recorded to assess side preferences (Deacon and Rawlins, 2006). The averages
and 95% confidence intervals of successful alternations were calculated for each group.
Statistical significance of dichotomous alternation behavior between groups was
calculated by Fisher’s exact test. For side preferences, chi-square scores were
calculated with an expected arm entry of 50% for left and right sides.
Novel Object Recognition
The novel object recognition (NOR) test consisted of three sequential phases: (1)
habituation, (2) sample-object training, and (3) novel-object exploration. All sessions
were recorded with a digital video camera. Mice were habituated to a black-walled,
grey-floored acrylic 25 × 25 cm
2
container for 5 min. In the sample-object phase, mice
were allowed to explore the container with two identical objects for 5 min. After a
retention interval of 90 min, one of the sample objects was replaced with a novel object
of similar volume but different color and shape. In the novel-object phase, mice were
allowed to explore both novel and familiar objects for 3 min. Object recognition was be
calculated for each mouse as percent novelty exploration index (% NEI) with the
following formula: [N / (N + F)] × 100%, with N and F indicating the time spent exploring
the novel and familiar objects in seconds, respectively (Bevins and Besheer, 2006;
87
Bortolato et al., 2010; Clinton et al., 2007). % NEI was not calculated for mice that did
not explore either objects.
Prepulse Inhibition
The SR-LAB Startle Response System (San Diego Instruments, San Diego, CA)
measured prepulse inhibition (PPI) in mice, according to optimal parameters in Basic
Protocol 2 for mice with 129Sv background (Geyer and Dulawa, 2003). The sound-
attenuating chamber consisted of a plastic cylinder mounted on a piezoelectric motion
sensor that detects movement; a computer measured the startle reflex in response to
acoustic stimuli. Sound levels were calibrated and measured with a CEM DT-85A sound
level meter.
Mice were acclimated to the chamber for 5 min with a background noise of 70 dB.
The acoustic startle reflex was measured in response to a 120-dB, 40-ms burst.
Prepulse inhibition was measured with prepulse stimuli of 3, 6, and 12 dB above
background. The % PPI for each prepulse intensity was calculated as 100% × [(pulse-
alone) - (prepulse + pulse score)] / pulse-alone score.
Carotid Blood Flow
Prior to necropsy, mice were anesthetized with isoflurane. Doppler ultrasound
echographs of the right common carotid artery were assessed noninvasively using a
Vivid 7 Dimension ultrasound system with a 6–13 MHz linear transducer (GE
Healthcare). Echographs were analyzed by measuring the time-averaged mean
(TAMAX) velocity and maximum vessel diameter (d) using the supplied EchoPAC
software. Carotid blood flow in mL/min was calculated as TAMAX × π × (d / 2)
2
.
88
Tissue Collection
Brain pathology was evaluated at 13 months of age. At necropsy, mice were
anesthetized with ketamine plus xylazine (100 mg/kg + 10 mg/kg i.p.). Following
thoracotomy, blood was harvested through the right atrium into an EDTA-coated tube.
Mice were transcardially perfused with PBS at a rate of 6 mL/min until the liver
blanched. Following perfusion, the brain was removed and sliced into equal
hemispheres.
The right hemisphere was immediately dissected into subregions: anterior cortex,
hippocampus, and amygdala. Subregions were flash frozen in the vapor phase of liquid
nitrogen and stored at -80 °C.
The left hemisphere was immersion-fixed in 4% paraformaldehyde at 4 °C for 48 hr.
Following fixation, the tissue was transferred to 20% sucrose at 4 °C for cryoprotection.
After the tissue sunk, it was quickly frozen in isopentane cooled with crushed dry ice.
The tissue was stored at -80 °C until sectioning. Tissues were sectioned coronally in 30
µm increments on a cryostat.
EPC Flow Cytometry
Following blood collection, 100 µL of whole blood was transferred to polypropylene
culture tubes and incubated with 1 µg of FITC rat anti-mouse Ly-6A/E (Sca-1) and 1 µg
of PE Rat Anti-Mouse Flk-1 antibodies (BD Biosciences) for 30 min at 4 °C in the dark.
RBCs were lysed with RBC Lysis Buffer (eBioscience). Cells were washed with PBS
and fixed with 2% formaldehyde until flow cytometric analysis.
89
Circulating EPCs were designated as Flk1+Sca1+. Flk1, fetal liver kinase 1, is an
endothelial cell marker. Flk1 is also known as kinase insert domain receptor (KDR) and
VEGFR-2. Sca1, stem cell antigen 1, is a hematopoietic stem cell marker. Cells were
gated first on Flk1, then on Sca1. This represented the percentage of total nucleated
cells in the endothelial lineage (Flk1) that were progenitors (Sca1). Sca1+ progenitor
cells were also analyzed.

Figure 9 EPC Gating Methodology
Cells were gated first on Flk1, then Sca1. Double-positive Flk1+Sca1+ cells were
considered EPCs.

Hippocampal Aβ Immunoassay
Soluble and insoluble Aβ concentrations were quantified by electrochemiluminescent
detection using MSD V-PLEX Aβ Peptide Panel 1 (6E10) Kits (Meso Scale Discovery,
US). Flash frozen unilateral hippocampi were homogenized in Neuronal Protein
Extraction Reagent (Thermo Scientific, MA) with 1× Halt Protease Inhibitor Cocktail and
EDTA in a Bullet Blender using 0.5 mm glass beads (Next Advance, NY). Homogenates
were centrifuged at 22,000 × g at 4 °C for 20 min. The supernatant was transferred to a
90
new tube, representing the soluble fraction. Next, 50 µL of 70% formic acid was added
to the remaining pellet, homogenized, and neutralized 1:14 by adding 750 µL of 1 M Tris
HCl. The acid-soluble homogenate was centrifuged at 22,000 × g for 20 min. The acid-
soluble supernatant was transferred to a new tube, representing the insoluble fraction.
The protein concentrations of the soluble and insoluble fractions were measured
using the Pierce BCA Protein Assay Kit (Thermo Scientific, MA) and absorbance at 280
nm. Insoluble fractions were normalized to total soluble protein. In the MSD plates, 25
µg of total protein was loaded into each well and processed according to manufacturer
instructions. Standard curves were established: Aβ38, 0–3000 pg/mL; Aβ40, 0–10,000
pg/mL; Aβ42, 0–3000 pg/mL.
Statistical Analysis
Statistical analyses were conducted with GraphPad Prism 6. Ordinary one-way
analysis of variance (ANOVA) was used to calculate statistical significance of groups (α
= 0.05) compared to appropriate vehicle controls. Multiple comparisons were corrected
by using Dunnett’s test. Outliers were identified and excluded from analysis using the
ROUT method (Q = 1%). Two-way ANOVA was used to assess whether genotype or
treatment had any effect on behavioral tests, carotid blood flow, and circulating EPCs.
Multiple comparisons were corrected by using Dunnett’s test.
91
Results
Working Memory Tests: T-maze & Novel Object Recognition
Behavioral tests were conducted in all mice, including all sexes, genotypes, and
treatments. These tests included spontaneous alternations in the T-maze at 9–10
months of age, NOR test at 11 months of age, and prepulse inhibition (PPI) at 12
months of age.
A total of 9 spontaneous alternation trials were conducted for each mouse in the T-
maze. No statistically significant differences in spontaneous alternations were observed
between genotypes or treatment effects. No statistically significant side preferences
were observed between all groups (degrees of freedom [df] = 15).
No significant difference was observed in % NEI between vehicle-treated 3xTg-AD
mice and NonTg mice, or between treated 3xTg-AD mice. A significant increase in %
NEI was observed in the Ang-(1–7)-treated and candesartan-treated groups compared
to vehicle-treated NonTg controls (p < 0.05 and p < 0.01, respectively).
92

Figure 10 Spontaneous alternations in T-maze.
Data were reported as the mean and 95% confidence interval of successful alternations.
No significant differences in alternation behavior were observed when compared to
vehicle-treated controls as determined by Fisher’s exact test.

Male
Female
0
20
40
60
80
100
% Alternation
Tg vehicle
NonTg vehicle
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
Male
Female
0
20
40
60
80
100
% Alternation
NonTg vehicle
NonTg A(1-7)
NonTg CAND
NonTg A(1-7) + CAND
93
Vehicle A(1-7) CAND A(1-7) + CAND
Male 3xTg-AD 60 (4.5) 54 (0.6) 46 (0.6) 48 (0.3)
Male NonTg 48 (0.3) 54 (0.7) 56 (1.3) 66 (12.7)
Female 3xTg-
AD
46 (0.6) 53 (0.3) 58 (2.0) 54 (0.6)
Female NonTg 50 (0.0) 63 (6.4) 60 (3.6) 44 (1.1)

Table 3 Mice exhibited no side preferences on the T-maze.
Data were reported as % right-arm entries (chi-square score). Chi-square scores were
calculated from the observed entries vs. an expected arm entry of 50% for each side.
No groups showed a statistically significant preference for either side (df = 15).

Figure 11 Novel Exploration Indices.
% Novel exploration index (% NEI) was calculated as the percentage of time spent
exploring the novel object over total exploration time. Data were reported as mean ±
SEM. Ang-(1–7) and CAND groups increased the % NEI compared to vehicle-treated
female NonTg mice (* p < 0.05, ** p < 0.01).
NonTg vehicle
Tg vehicle
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
25
50
75
100
NEI %
Male
NonTg vehicle
Tg vehicle
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
25
50
75
100
NEI %
Female
Male
Female
0
25
50
75
100
NEI %
NonTg
* **
vehicle
A(1-7)
CAND
A(1-7) + CAND
94

Figure 12 Total Exploration Time.
Total exploration (s) between both novel and familiar objects was calculated during the
novel-object exploration phase. Data were reported as mean ± SEM. Total exploration
time did not differ significantly amongst groups.

Prepulse Inhibition
No significant differences were observed between treatment groups in the PPI of
3xTg-AD mice and NonTg controls at prepulse levels of 3, 6, or 12 dB as determined by
two-way ANOVA. Genotype-specific differences were observed in acoustic startle reflex
and PPI. NonTg controls exhibited significantly diminished acoustic startle reflex (p <
0.001 in males, p < 0.01 in females) and PPI compared to 3xTg-AD mice.
NonTg vehicle
Tg vehicle
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
50
100
150
Total Exploration (s)
Male
NonTg vehicle
Tg vehicle
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
50
100
150
Total Exploration (s)
Female
Male
Female
0
50
100
150
Total Exploration (s)
NonTg
vehicle
A(1-7)
CAND
A(1-7) + CAND
95

Figure 13 Acoustic Startle.

Figure 14 Prepulse Inhibition.
NonTg vehicle
3xTg-AD vehicle
3xTg-AD A(1-7)
3xTg-AD CAND
3xTg-AD A(1-7) + CAND
0
50
100
150
200
250
3xTg-AD Male Startle
Startle magnitude
***
NonTg vehicle
3xTg-AD vehicle
3xTg-AD A(1-7)
3xTg-AD CAND
3xTg-AD A(1-7) + CAND
0
50
100
150
200
250
3xTg-AD Female Startle
Startle magnitude
**
NonTg vehicle
3xTg-AD vehicle
3xTg-AD A(1-7)
3xTg-AD CAND
3xTg-AD A(1-7) + CAND
0
20
40
60
80
100
% PPI
3xTg-AD Male PPI
vehicle
A(1-7)
CAND
A(1-7) + CAND
0
20
40
60
80
100
NonTg Male PPI
% PPI
NonTg vehicle
3xTg-AD vehicle
3xTg-AD A(1-7)
3xTg-AD CAND
3xTg-AD A(1-7) + CAND
0
20
40
60
80
100
3xTg-AD Female PPI
% PPI
vehicle
A(1-7)
CAND
A(1-7) + CAND
0
20
40
60
80
100
NonTg Female PPI
% PPI
3 dB
6 dB
12 dB
96
Candesartan reduced carotid blood flow, except in NonTg female mice
There was no statistically significant difference amongst sex or genotype in carotid
blood flow in vehicle-treated mice (p > 0.05). However, there were sex-specific
differences depending on the genotype.
Genotype accounted for 27.02% of the total variation in carotid blood flow in female
mice in a two-way ANOVA assessing the effects of treatment and genotype (p < 0.01).
Candesartan reduced carotid blood flow in female mice in a genotype-dependent
manner. Candesartan reduced carotid blood flow in female 3xTg-AD mice (p < 0.05),
while this effect was not observed in female NonTg mice.
Genotype had no statistically significant effect on carotid blood flow in male mice.
Candesartan reduced carotid blood flow in both male 3xTg-AD and NonTg mice.

Figure 15 Carotid Blood Flow.
Carotid blood flow was measured by Doppler ultrasound as a surrogate measure of
cerebral blood flow. Candesartan significantly reduced carotid blood flow in male mice (*
p < 0.05).

Male
Female
0
5
10
15
Carotid blood flow by sex and genotype
Flow Rate (mL/min)
Veh
A(1-7)
CAND
A(1-7) + CAND
0
5
10
15
20
Flow Rate (mL/min)
Carotid blood flow in male mice
*
Veh
A(1-7)
CAND
A(1-7) + CAND
0
5
10
15
20
Flow Rate (mL/min)
Carotid blood flow in female mice
3xTg-AD
NonTg
*
97
Candesartan increased circulating Flk1+Sca1+ EPCs
Candesartan increased circulating EPCs (% cells Flk1+Sca1+) in both male 3xTg-
AD and NonTg mice measured by flow cytometry in whole blood (p < 0.0001). There
was no difference in EPCs between 3xTg-AD and NonTg mice. Combination of Ang-(1–
7) and candesartan increased circulating Sca1+ cells in both male 3xTg-AD mice and
NonTg mice compared to vehicle-treated controls (p < 0.001). There was a trend
towards increased Sca1+ cells with candesartan treatment, but this was not statistically
significant. There was no difference in Sca1+ cells between vehicle-treated 3xTg-AD
and NonTg mice.

Figure 16 Candesartan increased circulating Flk1+Sca1+ EPCs.
Data were reported as mean ± SEM. Candesartan significantly increased circulating
Flk1+Sca1+ cells compared to vehicle-treated controls (**** p < 0.0001 for treatment
effect). Combination of Ang-(1–7) and candesartan increased circulating Sca1+ cells
compared to vehicle-treated controls (*** p < 0.001).

Veh
A(1-7)
CAND
A(1-7) + CAND
0.0
0.2
0.4
0.6
0.8
1.0
Circulating EPCs in male mice
% cells Flk1+Sca1+
****
Veh
A(1-7)
CAND
A(1-7) + CAND
0
2
4
6
8
Sca1+ Cells
% cells Sca1+
3xTg-AD
NonTg
***
98
Candesartan decreased hippocampal insoluble Aβ species
Candesartan decreased insoluble Aβ levels in dissected unilateral hippocampi of
male 3xTg-AD mice compared to vehicle-treated controls, with statistical significance in
insoluble Aβ38 levels (p < 0.05). Combination of Ang-(1–7) and candesartan further
reduced insoluble Aβ38 levels (p < 0.01).
Soluble Aβ40 species were elevated with candesartan and combination treatment
groups (p < 0.05).
Combination of Ang-(1–7) and candesartan ratio showed the greatest reduction in
the insoluble-to-soluble Aβ40 ratio (p < 0.05).
99

Figure 17 Candesartan decreased hippocampal insoluble Aβ species.
Candesartan reduced insoluble Aβ38 species in 3xTg-AD mice (p < 0.05). Combination
of Ang-(1–7) and candesartan also reduced insoluble Aβ38 species (p < 0.01) and
decreased the ratio of insoluble-to-soluble Aβ40 (p < 0.05). Candesartan and
combination groups increased soluble Aβ40 species (p < 0.05).

Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
20
40
60
80
100
Insoluble Aβ38 (pg/mL)
Insoluble Aβ38
* **
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
2
4
6
8
10
Soluble Aβ38 (pg/mL)
Soluble Aβ38
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
5
10
15
20
25
Insoluble:Soluble Aβ38 Ratio
Aβ38 Ratio
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
200
400
600
Insoluble Aβ40 (pg/mL)
Insoluble Aβ40
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
10
20
30
40
Soluble Aβ40 (pg/mL)
Soluble Aβ40
* *
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
10
20
30
40
50
Insoluble:Soluble Aβ40 Ratio
Aβ40 Ratio
*
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
500
1000
1500
Insoluble Aβ42 (pg/mL)
Insoluble Aβ42
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
2
4
6
8
10
Soluble Aβ42 (pg/mL)
Soluble Aβ42
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
50
100
150
Insoluble:Soluble Aβ42 Ratio
Aβ42 Ratio
100
Candesartan decreased survival of female 3xTg-AD mice, rescued by Ang-(1–7)
Candesartan decreased the survival of female 3xTg-AD mice treated with daily
subcutaneous injections at 1 mg/kg. This effect was abolished in the combination
candesartan and Ang-(1–7) group. Prior to death, female 3xTg-AD mice in the
candesartan group exhibited signs of lethargy and decreased perfusion to the
extremities, which was not evident in the combination group or NonTg controls. This
survival effect was not observed in male mice for the same dose. This may be related to
the observation that female 3xTg-AD mice had significantly lower carotid blood flow with
candesartan treatment, compared to male mice.

Figure 18 Survival Curve of Female 3xTg-AD Mice.
Daily subcutaneous injections of candesartan at 1 mg/kg decreased the survival of
female 3xTg-AD mice over an 8-month timeframe. This effect was rescued by
combination with Ang-(1–7) treatment at 500 µg/kg.
0 50 100 150 200
0
50
100
Days of Treatment
% Survival
Survival of female 3xTg-AD mice
Vehicle
Ang-(1-7)
CAND
Ang-(1-7) + CAND
101
Discussion
Behavioral Tasks
Behavioral tasks were attempted to assess the effects of treatment on working
memory and sensorimotor gating. No differences were observed between 3xTg-AD and
their NonTg controls on spontaneous alternations in T-maze or the novel object
recognition test. This may be due to the suboptimal validation of behavioral paradigms,
limited number of testing sessions, or lack of detectable deficits at 13 months of age. All
animals were subjected to behavioral testing, including all genotypes and treatments
(including treated NonTg groups), which limited the number of trials and statistical
power of certain tests such as the T-maze.
No difference was detected on spontaneous alternations or side preferences in the
T-maze between genotypes or treatments. There was a slight trend in which female
3xTg-AD mice performed better than their NonTg controls, but this was not statistically
significant. These results may be due to the limited number of trials per mouse. Data
collection was limited to only 9 alternation trials per mouse due to experimental
constraints. Since spontaneous alternation is a dichotomous choice (i.e. yes or no),
more trials may be more appropriate to achieve a significant statistical power. Further,
alternations were not matched to individual mice, which made ANOVA statistical
analysis impossible. Instead, Fisher’s exact test was conducted to compare differences
in proportions between groups, which may have limited the statistical power compared
to an ANOVA test.
102
No difference was detected on the NOR test between vehicle-treated 3xTg-AD and
NonTg mice. Only Ang-(1–7)-treated and candesartan-treated female NonTg mice
showed increased % NEI with treatment compared to vehicle, but this was likely due to
the unusual finding that the vehicle-treated female NonTg mice appeared to show
preference for the familiar object (i.e. % NEI < 50%). The NOR test differs based on
background strains, and background strains have been known to carry distinct
behavioral phenotypes (Stover et al., 2015). It is possible that the mice could not
distinguish selected familiar and novel reliably. Future studies may consider object
validation prior to NOR testing. Further, habituation (5 min) and retention times (90 min)
may have not been appropriate. While total exploration did not differ significantly
between groups, within-group variations in exploration were high and some mice
displayed limited exploration, which is possibly indicative of poor acclimation. Longer
times or multiple habituation sessions may be beneficial for mice to acclimate to the
testing environment.
Interestingly, genotype-specific PPI effects were observed in the 3xTg-AD model, in
which their NonTg controls exhibit significantly diminished acoustic startle reflex and
PPI. While treatments did not significantly affect PPI in either genotype, candesartan
appeared to show a modest improvement in the PPI of male NonTg mice, while the
opposite was observed in female NonTg mice. Similar deficits in acoustic startle
response and PPI were observed in the NonTg controls in another study investigating
the effects of physical exercise in 3xTg-AD mice (García-Mesa et al., 2011).
103
Although PPI tasks translate readily from rodents to humans in diseases such as
schizophrenia, the applicability of PPI in AD remains questionable. While the
hippocampus and entorhinal cortex are both affected in AD and involved in PPI, clinical
studies have found conflicting results in PPI deficits between AD/MCI patients and
healthy controls (Hejl et al., 2004; UEKI et al., 2006). In the first study, no deficits were
observed in either MCI or AD cases. Further, PPI was not found to correlate with
cognitive performance or neuropsychiatric symptoms (Hejl et al., 2004). In a separate
study, enhanced PPI was observed in patients with amnestic MCI, while those with mild
AD had significantly less PPI than controls (UEKI et al., 2006). Ultimately, there is
insufficient evidence for either PPI enhancement or deficits in cognitive decline, and PPI
testing is not routinely used in the clinical diagnosis of memory loss, MCI, or AD.
With these behavioral results in consideration, no cognitive deficit was evident
between genotypes at 13 months of age as assessed by spontaneous alternations on T-
maze and the NOR test. Without distinct cognitive deficits between vehicle-treated
3xTg-AD mice and their NonTg controls, it was uncertain whether treatment elicited any
behavioral changes.
The chosen behavioral paradigms may not be able to detect changes in cognitive
deficits in 3xTg-AD mice at this age. A 2015 study investigated the cognitive
performance of male and female 6.5-month-old 3xTg-AD mice on spontaneous
alternations in Y-maze, NOR test, and Barnes maze (Stover et al., 2015). Interestingly,
this validation study found nearly the same findings: 3xTg-AD mice performed better on
the spontaneous alternation task, and no difference in novel object recognition. The
104
3xTg-AD mice exhibited slight, but statistically significant deficits in the Barnes maze,
indicating mild yet detectable deficits in spatial learning and memory. Behavioral tasks
involving spatial memory such as the Barnes maze or Morris Water Maze may be more
appropriate to detect cognitive deficits in 3xTg-AD mice with sensitivity.
Vascular Mechanisms
Vascular repair mechanisms, such as EPCs, are known to be diminished in AD
patients (Lee et al., 2010). Further, the progeny of EPC colonies from AD patients
exhibit dysfunctional characteristics such as higher senescence-associated β-
galactosidase activity (Lee et al., 2010). We surmised that candesartan-induced AT1
blockade may improve endothelial function by increasing circulating EPCs, thereby
restoring cerebral blood flow and neurovascular coupling.
Carotid Blood and Survival
I hypothesized that candesartan may restore neurovascular coupling in 3xTg-AD
mice by increasing cerebral blood flow. We measured carotid blood flow by Doppler
ultrasound as a surrogate measure of cerebral blood flow, as blood flow through the
carotid arteries is the major route that supplies oxygenated blood to the brain.
Contrary to our hypothesis, candesartan significantly reduced carotid blood flow in
both male and female 3xTg-AD mice. This reduction in carotid blood flow may be
related to the undesired hypotensive effects of AT1 blockade. We did not measure blood
pressure in response to long-term treatment of candesartan. However, it was previously
been reported that subcutaneous administration of candesartan at the same dose, 1
105
mg/kg/day, did not influence systolic blood pressure after 4 weeks of treatment in
diabetic db/db mice and their heterozygous controls (Callera et al., 2016).
As shown above, there were unexpected deaths in the female transgenic mice
treated with candesartan. Prior to death, candesartan-treated female 3xTg-AD mice
exhibited signs of lethargy and decreased perfusion to the extremities, which points
towards potential hypotensive effects. Undesired hypotensive effects may have been
responsible for the decreased survival in female 3xTg-AD mice treated with long-term
candesartan. Of note, this gender-selective adverse event has not been seen in
humans taking candesartan.
Due to the decreased survival of female 3xTg-AD mice in the candesartan group,
male 3xTg-AD mice were emphasized for pathology and mechanistic studies to avoid
survivor bias. It may be prudent to pursue alternative dosing in female 3xTg-AD mice for
future studies, as the current dose of 1 mg/kg candesartan subcutaneous treatment
may have induced undesired hypotensive effects.
EPCs
The results show that candesartan significantly increased circulating Flk1+Sca1+
EPCs in all mice. This effect may have been driven by the increase in Sca1+
progenitors, with the strongest effect in the combination group. Interestingly, while
candesartan increased Flk1+Sca1+ EPCs, the same effect was not observed when
coadministered with Ang-(1–7). Instead, there was a reduction of EPCs, which may
reflect either diminished mobilization of EPCs from the BM, or perhaps increased
differentiation or homing to sites of injury. The latter is more likely, considering the
106
significant increase in circulating Sca1+ progenitors with combination treatment.
Furthermore, our lab has also shown that Ang-(1–7) increased circulating Flk1+Sca1+
EPCs in a diabetic mouse model (Papinska et al., 2015).
However, circulating EPCs did not differ between vehicle-treated 3xTg-AD mice and
their NonTg controls. This suggests that EPC deficits may not be relevant in transgenic
animal models associated with early-onset familial AD mutations. Clinical observations
of EPC deficit in AD may differ from those seen in transgenic animal models, as most
clinical AD cases comprise of the late-onset, sporadic type. Based on this evidence, it is
unlikely that the reduction of Aβ was attributed to increased EPC number. It is possible
that EPCs may be downregulated with older age and/or higher disease severity.
However, it is also important to note that EPC markers do not readily translate between
mice and humans, and EPC numbers may differ as a result.
Reduction of Hippocampal Aβ
The reduction of hippocampal Aβ species in 3xTg-AD mice treated with candesartan
and Ang-(1–7) points towards RAS modulation as a promising strategy for preventing
Aβ deposition. Daily subcutaneous injections of candesartan for 8 months significantly
reduced insoluble Aβ38 levels in the hippocampi of male 3xTg-AD mice compared to
vehicle-treated controls. This effect was further pronounced in combination with Ang-(1–
7), which had the highest degree of Aβ38 reduction and a significant reduction in
insoluble-to-soluble Aβ40 ratio. While Ang-(1–7) alone did not significantly reduce Aβ
levels, a downward trend in insoluble Aβ levels was observed.
107
Interestingly, soluble Aβ40 levels were significantly elevated in the candesartan and
combination group compared to the vehicle-treated controls. However, soluble Aβ levels
were orders of magnitude lower than their respective insoluble Aβ fraction (5–20 pg/mL
soluble vs. 500–2000 pg/mL insoluble protein). Thus, most of the detected Aβ was
present in the insoluble pool. Similarly, the insoluble-to-soluble Aβ ratio has been
reported to increase with AD progression, likely indicative of Aβ aggregation and plaque
formation. The decrease in the ratio of insoluble-to-soluble Aβ40 with RAS therapy may
reflect an increased clearance of soluble Aβ monomers prior to aggregation and plaque
formation.
It is important to note that treatment began at 5 months of age in 3xTg-AD mice,
prior to significant extracellular Aβ deposition that typically occurs at 9–12 months of
age (Oddo et al., 2003). Furthermore, combination therapy demonstrated the greatest
overall reduction in insoluble Aβ species. Taken together, these data suggest that
pharmacological intervention targeting the protective RAS may be beneficial in
preventing Aβ deposition.
Surmounting clinical evidence suggest that ARBs may reduce AD pathology and
delay the onset of dementia, which has spurred further investigation into the
mechanisms underlying this observation (Davies et al., 2011; Hajjar, 2012; Li et al.,
2010). Preclinical studies have shown conflicting results regarding ARBs in transgenic
AD models, likely due to the varied selection of ARBs, their degree of BBB penetration,
and delivery route (Corbett et al., 2012). Losartan, irbesartan, telmisartan, and
candesartan have been shown to attenuate the centrally mediated effects of Ang II in a
108
dose-dependent manner (Culman et al., 2002). The reduction in Aβ may also be due in
part to candesartan’s ability to pass through the BBB and act as a central AT1
antagonist and PPAR-γ agonist. Activation of PPAR-γ has been known to modulate
microglial activation, potentially increasing glial Aβ phagocytosis (Mandrekar-Colucci et
al., 2012).
At present, only one group has examined ARBs (valsartan, eprosartan) in male
3xTg-AD mice, and found no changes in cognition or amyloid deposition (Ferrington et
al., 2012; 2011). It is important to note these studies administered valsartan and
eprosartan in drinking water ad libitum for 2 or 6 months, and began treatment at 3–4
months or 9–10 months of age. The selection of ARBs that do not penetrate the BBB
effectively may have impaired the ability to effect central Aβ deposition.
AT1 Blockade vs. Mas Stimulation
Whether the Aβ-lowering effects of candesartan are attributed to AT1 blockade, AT2
activation, Mas activation, or a combination, is unknown. Central blockade of AT1
receptors reduces the pathological effects of Ang II through the AT1 receptor, and also
leads to more available Ang II to either: (1) bind to the AT2 receptor or (2) process
further to Ang-(1–7), which binds to the Mas receptor. Activation of the protective arm of
RAS, mediated by AT2 and Mas receptors, opposes the actions of pathological AT1
signaling. Mas activation induces vasodilatory, anti-inflammatory, and anti-oxidative
effects. While there is no direct plausible link to Aβ reduction, Mas activation is known to
facilitate hippocampal LTP, a proposed cellular mechanism underlying learning and
memory (Hellner et al., 2005). Based on the observation that combination therapy
109
demonstrated the greatest Aβ reduction, there may be a synergistic effect of AT1
blockade, AT2 activation, and Mas activation acting in concert.
A prior study showed that candesartan ameliorated SCO-induced cognitive
impairment, a model of cholinergic deficit (Tota et al., 2012). This effect was blunted by
coadministration with PD123,319, an AT2 antagonist, which suggested AT2 may be
beneficial in protecting cholinergic function. It was postulated that AT2 activation may
elicit beneficial effects in AD by increasing cerebral blood flow through the vasodilatory
effects of bradykinin and nitric oxide release (Mogi and Horiuchi, 2013). However, no
studies to date have looked at AT2 function in transgenic AD models based on Aβ
deposition. Future mechanistic studies may shed light on the specific RAS pathways
responsible for Aβ reduction by utilizing combinations of centrally active AT1, AT2, and
Mas agonists along with specific antagonists such as an ARB, PD123,319, and A-779,
respectively.
Candesartan has been known to activate other parts of the RAS, specifically
upregulating ACE2 and Mas expression (Callera et al., 2016). Decreased plasma ACE2
activity has been observed in AD patients, and ACE2 has recently been found to
degrade Aβ43 species, a particularly amyloidogenic species responsible for early
plaque formation (Liu et al., 2014). Further mechanistic studies are warranted to
determine the causative pathways and their relative contributions underlying RAS-
mediated Aβ reductions in 3xTg-AD.
Alternatively, AT1 activation is known to impair neurovascular coupling and drive
vascular senescence (Kazama et al., 2004). The neurovascular hypothesis posits that
110
Aβ clearance across the BBB may be impaired due to either aberrant angiogenesis or
endothelial senescence, resulting in diminished efflux of Aβ through LRP1 transporters
(Zlokovic, 2005). Endothelial dysfunction resulting from chronic AT1 activation may
impair the ability of endothelium-resident efflux transporters to clear Aβ across the BBB.
Further, senescent endothelial cells impair functional hyperemia and cerebral blood flow
(Girouard and Iadecola, 2006).
Summary
Significant clinical evidence has shown that ARBs delay the progression of
dementia, which has spurred the investigation into the underlying mechanisms of ARB’s
effects in AD. Here we have shown for the first time that long-term treatment with
candesartan, a potent ARB, significantly decreased insoluble Aβ species in the 3xTg-AD
mouse model. Combination therapy with Ang-(1–7) showed the greatest overall
reduction in Aβ, providing further evidence that RAS modulation through inhibition of the
pathological AT1 signaling and activation of the protective RAS may be beneficial in AD.
As treatment began at 5 months of age prior to significant extracellular Aβ deposition,
we propose that RAS modulation may be a worthwhile therapeutic intervention in
preventing or delaying the progression of AD.
Taken together, these data suggest that Aβ reduction resulting from long-term
candesartan treatment was not associated with vascular mechanisms related to EPCs
or cerebral blood flow. Rather, Aβ reduction may have occurred through other Aβ
clearance mechanisms, such as catabolism by Aβ-degrading enzymes or degradation
by glial cells. Candesartan’s role as a partial PPAR-γ agonist may also modulate glial
111
activation. Further investigations into the mechanisms underlying Aβ reduction in 3xTg-
AD mice with RAS treatment are necessary.
It is unclear whether there is any cognitive benefit, as behavioral testing showed no
difference in working memory tests between 3xTg-AD mice and NonTg controls.
Behavioral tasks were limited in their scope, statistical power, and feasibility. Future
behavioral studies should focus on robust, sensitive, and validated measures of
cognition, such as spatial memory tasks, e.g. the Morris Water Maze or Barnes maze.
112
Chapter 3: Aβ Clearance Mechanisms
Introduction
The net balance of Aβ can be viewed as the sum of Aβ production and clearance
pathways (Santos et al., 2011). Aβ production depends on the amyloidogenic
processing of APP to generate Aβ. Aβ clearance can be mediated through several
mechanisms, including Aβ-degrading enzymes and glial phagocytosis. Clinical studies
have shown that ARBs reduce amyloid deposition markers, which may be mediated by
increased Aβ clearance (Hajjar, 2012).
The data from Chapter 2 indicated that 8-month treatment of candesartan and Ang-
(1–7) significantly reduced insoluble hippocampal Aβ levels in male 3xTg-AD mice.
Several possible underlying mechanisms may be associated with RAS treatment on Aβ
clearance. Systemic administration of RAS agents may be altering the brain RAS,
activating PPAR-γ pathways, and upregulating Aβ-degrading enzymes to promote Aβ
degradation. Beyond reducing neuroinflammation, PPAR-γ activation can also enhance
glial clearance of Aβ. I hypothesize that RAS treatment increased Aβ clearance in 3xTg-
AD mice by upregulating Aβ-degrading enzymes and increasing glial phagocytosis of
Aβ.
The 3xTg-AD mouse model relies on transgenic expression of mutant human genes
(APP
Swe
and PS1
M146V
) that increase Aβ deposition through the amyloidogenic pathway
(Van Dam and De Deyn, 2006). In the amyloidogenic pathway, APP is sequentially
cleaved by β- and γ-secretases to produce Aβ and APP terminal fragments (sAPPβ and
CTFβ). Bace1 and Psen1 encode for β-secretase and the presenilin-1 catalytic subunit
113
for γ-secretase, respectively. Transgenic expression of mutant App and Psen1 genes
result in constitutively high Aβ production through the amyloidogenic pathway. Central
infusion of losartan did not affect App, Bace1, and Psen1 expression in rats (Zhu et al.,
2011). Thus, I hypothesize that treatment did not affect amyloidogenic processing of
APP, but this must be confirmed experimentally.
Aβ-degrading enzymes in the brain can catabolize Aβ, including: IDE, NEP, ECE,
and ACE (Santos et al., 2011). IDE is the most abundant enzyme, which degrades
monomeric Aβ species. Neprilysin can also degrade Aβ plaques in addition to
monomers. To a lesser degree, ECE and ACE can also participate in Aβ degradation.
Coincidentally, some RAS enzymes catabolize both angiotensins and Aβ including:
ACE, neprilysin, and potentially ACE2. Therefore, RAS agents may modulate elements
of the brain RAS that have dual actions as Aβ-degrading enzymes.
Other centrally modulated RAS elements may also be involved in cognition,
particularly those implicated in learning and memory. Studies have shown that the
ACE2/Ang-(1–7)/Mas and Ang IV/AT4 pathways facilitate LTP, a major cellular
mechanism that underlies learning and memory (Hellner et al., 2005; Wayner et al.,
2001; Wright et al., 2013). While the true identity of the AT4 receptor is unknown, IRAP
is a potential candidate, encoded by the Lnpep (leucyl/cystinyl aminopeptidase) gene.
(Wright and Harding, 2011). As a related RAS aim, I characterized the expression of
central RAS elements that may be involved in learning & memory.
Studies have shown that PPAR-γ activation enhances glial amyloid clearance
through activation of LXRα target genes ApoE and ABCA1 (Heneka et al., 2011;
114
Mandrekar-Colucci and Landreth, 2010). Lipidation of ApoE by ABCA1 enhances the
Aβ-binding properties of ApoE (Wildsmith et al., 2013). Consequently, lipidated ApoE
binds Aβ to prevent aggregation and facilitate proteolytic degradation of Aβ by glia and
Aβ-degrading enzymes (Jiang et al., 2008). Candesartan and Ang-(1–7) both activate
the PPAR-γ pathway. I assessed glial activation through the PPAR-γ pathway as a
potential mechanism driving Aβ clearance.
These Aβ production and clearance mechanisms are summarized in the following
figure. To investigate these potential underlying mechanisms, mRNA gene expression
studies were conducted on genes involved in APP processing, Aβ-degrading enzymes,
glial activation, inflammation, PPAR-γ activation, and RAS enzymes and receptors.

115

Figure 19 Aβ Production and Clearance Pathways.
(a) APP processing pathways. The non-amyloidogenic pathway (left) involves
cleavage of APP by α- and γ-secretases to generate sAPPα and non-amyloidogenic
species. Conversely, the amyloidogenic pathway (right) involves sequential cleavage by
β- and γ-secretases to generate sAPPβ and pathological Aβ.
(b) Aβ Aggregation and Clearance. Extracellular Aβ monomers aggregate into fibrils
and plaques. Aβ-degrading enzymes can act upon these Aβ substrates, including:
insulin-degrading enzyme (IDE), neprilysin (NEP), endothelin-converting enzyme (ECE),
and angiotensin-converting enzyme (ACE). Microglia and astrocytes can also
phagocytose Aβ monomers, where intracellular Aβ-degrading enzymes can break them
down further. Figures for microglia and astrocytes were reproduced from Wikimedia
Commons under the Creative Commons licenses. These works were attributed to
Microsome and Cancer Research UK, respectively.
Aβ
APP
Aβ
β
Aβ
γ
sAPPβ sAPPα
α
γ
extracellular space
cytosol
membrane
amyloidogenic non-amyloidogenic
(a) APP Processing (b) Aβ Aggregation & Clearance
IDE
monomers fibrils plaques
NEP
ACE
ECE
IDE
NEP
ECE
NEP
microglia astrocyte
116
Methods
mRNA Expression
To investigate the underlying mechanisms of RAS treatment on Aβ clearance, mRNA
expression of cortical brain tissue was measured using quantitative reverse transcription
polymerase chain reaction (qRT-PCR). The expression of specific pathways was
evaluated, including: APP processing, Aβ-degrading enzymes, glial activation, PPAR-γ
activation, and RAS enzymes and receptors.
Total RNA was extracted from homogenized anterior cortical tissue from the right
brain hemisphere of mice using TRIzol reagent. RNA was treated with DNase using the
TURBO DNA-free Kit to remove contaminating DNA. cDNA was synthesized from the
RNA template using the RevertAid RT Reverse Transcription Kit (Thermo Fisher).
qPCR was performed using Maxima SYBR Green/ROX qPCR Master Mix reagents
in a 25 µL reaction with 0.4 µM primers on an Applied Biosystems 7300 Real Time PCR
System. Amplification conditions consisted of a pre-treatment step at 50 °C for 2 min,
initial denaturation at 95 °C for 10 min, and 40 cycles of denaturation at 95 °C for 15 sec
and annealing/extension at 60 °C for 1 min. All reactions were conducted in triplicate.
The relative expression of target genes (i.e. fold changes) were calculated using the
ΔΔCT method against Rps29 as an endogenous control. Rps29 encodes the 40S
ribosomal protein S29. Rps29 amplified early and displayed excellent, low coefficients
of variation (CVs) in all cDNA samples. mRNA fold changes were calculated relative to
vehicle-treated 3xTg-AD mice. A maximum allowable CT value of 40 was used to avoid
skewing of lowly expressed genes towards detectable samples. Primers were
117
developed in our lab or sourced from the PrimerBank database (Spandidos et al.,
2009). Primer sequences for App and Psen1 were designed to not overlap with the
transgenic mutation sequence.

118
Gene Synonyms Forward Reverse
Rps29 GCAAATACGGGCTGAACATG GACTAGCATGATCGGTTCCAC
APP Processing
App TCCGAGAGGTGTGCTCTGAA CCACATCCGCCGTAAAAGAATG
Bace1 GGAACCCATCTCGGCATCC TCCGATTCCTCGTCGGTCTC
Psen1 GGTGGCTGTTTTATGTCCCAA CAACCACACCATTGTTGAGGA
Aβ-degrading Enzymes
Ide CAGAAGGACCTCAAGAATGGGT GCCTCGTGGTCTCTCTTTATCT
Mme neprilysin (Nep) CTCTCTGTGCTTGTCTTGCTC GACGTTGCGTTTCAACCAGC
Ece1 TCTCCGAGGGCGATGTGTA CTTCTCCACCGAGGTCCGA
Ece2 AGTTACCGGAGAAGAACTTCCA CGCAACCTAGCACGAGGATAC
Ace ATCCGTAACCATCACAGCC ACTCAAACACCATGTCCCC
Glial Activation
Iba1 Aif1 ATCAACAAGCAATTCCTCGATGA CAGCATTCGCTTCAAGGACATA
Ptprc Cd45 ATGGTCCTCTGAATAAAGCCCA TCAGCACTATTGGTAGGCTCC
Gfap CGGAGACGCATCACCTCTG AGGGAGTGGAGGAGTCATTCG
PPAR-γ/LXR Activation
Pparg TTTTCAAGGGTGCCAGTTTC AATCCTTGGCCCTCTGAGAT
Apoe CTGACAGGATGCCTAGCCG CGCAGGTAATCCCAGAAGC
Abca1 AAAACCGCAGACATCCTTCAG CATACCGAAACTCGTTCACCC
Nr1h3 LXR-α CTCAATGCCTGATGTTTCTCCT TCCAACCCTATCCCTAAAGCAA
Renin-Angiotensin System Enzymes and Receptors
Agt Angiotensinogen GCAAAACTCAGTGCTGTCAC CTGTCACCCCAGTATCCAAAC
Ren1 Renin GTTCATCCTTTATCTCGGCTCC AGGTCAAAGGAAATGTCGGG
Ace ATCCGTAACCATCACAGCC ACTCAAACACCATGTCCCC
Ace2 CCCCAAAATGTGTCTGATGTC CTGGTAAGGTGGCTCAAGTG
Agtr1a AT1 AACAGCTTGGTGGTGATCGTC CATAGCGGTATAGACAGCCCA
Agtr2 AT2 CCGAGAACAGGAAGTCAAGTAG ATCTCATGCCAAAGGACCAG
Mas1 CGCTTCAGGGAGTCCTTAAAAG CATTCCACTGTTTCTGTCCAC
Mrgprd TTTTCAGTGACATTCCTCGCC GCACATAGACACAGAAGGGAGA
Lnpep AT4 TCTTACAGAACAAGTGCAGTGG TCGGTCTTTGTCACTCAGAACA
Table 4 PCR Primer Sequences.
Mouse-specific primers were designed to assess the relative mRNA expression of
genes relative to an endogenous control, Rps29. Gene expression was evaluated from
cortical brain tissue using quantitative reverse transcription PCR (qRT-PCR). Pathways
of interest included: APP processing, Aβ-degrading enzymes, glial activation, PPAR-
γ/LXR activation, and the renin-angiotensin system.
119
Soluble APPα/β Immunoassay
Soluble APP α and β were quantified by electrochemiluminescent detection using
sAPPα and Swedish sAPPβ Kits (Meso Scale Discovery, US). Flash frozen unilateral
hippocampi were homogenized in Neuronal Protein Extraction Reagent (Thermo
Scientific, MA) with 1× Halt Protease Inhibitor Cocktail and EDTA in a Bullet Blender
using 0.5 mm glass beads (Next Advance, NY). Homogenates were centrifuged at
22,000 × g at 4 °C for 20 min. The supernatant was transferred to a new tube,
representing the soluble protein fraction.
The protein concentrations of the soluble proteins were measured using the Pierce
BCA Protein Assay Kit (Thermo Scientific, MA) and absorbance at 280 nm. In the MSD
plates, 10 µg of total protein was loaded into each well and processed according to
manufacturer instructions. Standard curves were established for each analyte: 0–1000
ng/mL.
Statistical Analyses
Statistical analyses were conducted with GraphPad Prism 6. Ordinary one-way
analysis of variance (ANOVA) was used to calculate statistical significance of groups (α
= 0.05) compared to vehicle-treated 3xTg-AD mice. Multiple comparisons were
corrected by using Dunnett’s test. Outliers were identified and excluded from analysis
using the ROUT method (Q = 1%).
120
Results
The mRNA expression levels of genes in the cortex of 3xTg-AD mice were assessed
by qRT-PCR. Pathways of interest included: APP processing, Αβ-degrading enzymes,
glial activation, PPAR-γ/LXR activation, and the RAS.
APP Processing

Figure 20 Cortical mRNA expression of APP processing genes.
Candesartan alone and in combination with Ang-(1–7) increased App mRNA fold
change to 1.2, with statistical significance in the combination group (p < 0.05).
Candesartan alone and in combination with Ang-(1–7) also significantly modulated the
expression of Bace1 and Psen1 mRNA. Bace1 mRNA fold change was reduced to 0.5
in candesartan and combination groups (p < 0.001 and p < 0.0001, respectively). Psen1
mRNA fold change was increased to 3.0 and 3.8, respectively, in candesartan and
combination groups (both p < 0.0001).
No changes were observed in the mRNA expression of APP processing genes
between vehicle-treated NonTg and 3xTg-AD mice, or with Ang-(1–7) treatment in
3xTg-AD mice.

NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
App
*
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
mRNA Fold Change
Bace1
*** ****
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
1
2
3
4
5
mRNA Fold Change
Psen1
**** ****
121
Soluble APPα and APPβ

Figure 21 Hippocampal sAPPα and Swedish sAPPβ levels.
A significant difference in hippocampal sAPPα and Swedish (sw) sAPPβ levels were
observed between NonTg and 3xTg-AD mice. NonTg mice had a mean of 5.5 pg/mL
sAPPα and 14.8 pg/mL sw sAPPβ. In contrast, vehicle-treated 3xTg-AD mice had
significantly higher (~500–900× higher) levels in the nanogram range: 2.9 ng/mL sAPPα
and 12.8 ng/mL sw sAPPβ. 3xTg-AD mice treated with Ang-(1–7) or candesartan did
not show any significant differences in sAPPα or sAPPβ levels. The sw sAPPβ-to-
sAPPα ratios were approximately 5:1, and did not differ in 3xTg-AD mice across all
treatments.

NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
1
2
3
4
sAPPα
Concentration (ng/mL)
****
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
5
10
15
20
Swedish sAPPβ
Concentration (ng/mL)
****
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
2
4
6
sw sAPPβ/sAPPα ratio
sw sAPPβ/sAPPα ratio
122
Aβ-degrading Enzymes

Figure 22 Cortical mRNA expression of Aβ-degrading enzymes.
RAS therapy significantly upregulated the mRNA expression of monomeric Aβ-
degrading enzymes Ide and Ece2. Compared to vehicle, Ang-(1–7) increased Ide
mRNA fold change to 1.6 (p < 0.05), candesartan to 1.8 (p < 0.01), and the combination
of both Ang-(1–7) and candesartan to 2.0 (p < 0.001). Ece1 and Ece2 showed a similar
trend of upregulated expression. The combination group increased Ece2 mRNA fold
change to 2.3 (p < 0.05).
The mRNA expression levels of neprilysin and ACE were not significantly changed
by either Ang-(1–7) or candesartan. The combination group increased neprilysin mRNA
fold change to 1.5, but this was not statistically significant.
Tg Veh
Tg A(1-7)
Tg CAND
Tg CAND + A(1-7)
0
1
2
3
mRNA Fold Change
Ide
* ** ***
Tg Veh
Tg A(1-7)
Tg CAND
Tg CAND + A(1-7)
0
1
2
3
mRNA Fold Change
Ece1
Tg Veh
Tg A(1-7)
Tg CAND
Tg CAND + A(1-7)
0
1
2
3
mRNA Fold Change
Mme (NEP)
Tg Veh
Tg A(1-7)
Tg CAND
Tg CAND + A(1-7)
0
1
2
3
Ece2
mRNA Fold Change
*
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Ace
ACE mRNA Fold Change
123
Glial Activation

Figure 23 Cortical mRNA expression of glial activation markers.
Microglial activation was assessed by Iba1/CD45, and astrocytic activation by GFAP.

Candesartan alone and in combination with Ang-(1–7) significantly increased the
mRNA expression of microglial markers Iba1 and CD45. No significant changes were
observed with astrocyte marker Gfap.
Candesartan significantly increased Iba1 mRNA fold change to 1.4 (p < 0.01) and
CD45 mRNA fold change to 1.5 (p < 0.05). Combination treatment with candesartan and
Ang-(1–7) increased Iba1 mRNA fold change to 1.4 (p < 0.01) and CD45 mRNA fold
change to 1.6 (p < 0.01).
Gfap mRNA fold change increased to 1.8 with candesartan and 1.4 with both
candesartan and Ang-(1–7), but these changes were not significant. No trends or
significant changes were evident in Iba1, CD45, or Gfap mRNA between vehicle-treated
3xTg-AD mice and NonTg controls.

NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
Iba1
** **
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
Ptprc (CD45)
* **
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
1
2
3
mRNA Fold Change
Gfap
124
PPAR-γ/LXR Activation

Figure 24 Cortical mRNA expression of PPAR-γ and LXR target genes.
PPAR-γ and LXR activation genes were assessed to evaluate their role in Aβ
clearance through microglia and astrocytes. No differences were observed in ABCA1,
APOE, LXRα, or PPAR-γ mRNA between NonTg and 3xTg-AD mice. There was a slight
trend of increased ABCA1 mRNA fold change to 1.2 with all RAS treatments, but this
change was not significant. APOE and LXRα mRNA levels were unchanged. PPAR-γ
mRNA fold change was significantly increased to 1.4 with combination treatment of Ang-
(1–7) and candesartan (p < 0.05). Ang-(1–7) alone or candesartan alone showed no
changes in PPAR-γ mRNA levels.

NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Abca1
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Nr1h3 (LXRα)
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Apoe
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
Pparg
*
125
Renin-Angiotensin System

Figure 25 Cortical mRNA expression of RAS enzymes and receptors.
No significant difference was observed in angiotensinogen (Agt) mRNA fold changes
with treatments, although there was a slight, insignificant decrease to 0.7 in the
combination group. No significant changes in mRNA levels were observed in Agt, RAS
enzymes, or receptors between NonTg and 3xTg-AD mice.
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0
1
2
3
mRNA Fold Change
Agt
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
Ace2
**
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
mRNA Fold Change
Mas1
* **
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Ren1
mRNA Fold Change
*
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
2.0
Agtr1a
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.00001
0.0001
0.001
0.01
0.1
1
10
mRNA Fold Change
Mrgprd
*** ***
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Ace
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
0.0
0.5
1.0
1.5
Agtr2
mRNA Fold Change
NonTg Veh
Tg Veh
Tg A(1-7)
Tg CAND
Tg A(1-7) + CAND
1
10
100
1000
10000
mRNA Fold Change
Lnpep
**** ****
126
Ren1 mRNA levels exhibited high variance in within-sample replicates. Ren1 mRNA
fold changes for Ang-(1–7), candesartan, and the combination treatments were reduced
to 0.3, 0.3, and 0.9, respectively. Only Ang-(1–7) showed a significant reduction in Ren1
mRNA levels (p < 0.05.)
Ace mRNA levels were not affected by genotype or treatments. Candesartan
significantly upregulated Ace2 mRNA by a fold change of 1.4 (p < 0.01). No other
changes in ACE2 mRNA levels were observed. No changes in AT1 and AT2 mRNA
levels were observed with treatment.
Mas1 mRNA fold changes were significantly upregulated to 1.5 with Ang-(1–7) alone
(p < 0.05) and to 1.6 with candesartan alone (p < 0.01), respectively. While there was
an increase to 1.4 in Mas1 mRNA fold change in the combination group, this change
was not significant.
Mrgprd mRNA levels were significantly decreased by over 1000-fold with
candesartan and combination treatments (both p < 0.001). Lnpep mRNA levels were
also increased by over 1000-fold (both p < 0.0001). While the mean of Lnpep
expression in the vehicle-treated NonTg group was near 1000-fold, this change was not
statistically significant due to the large variance of Lnpep expression in individual
animals.
127
Discussion
Candesartan modulated APP processing genes Bace1 and Psen1
Unexpectedly, candesartan treatment affected APP processing genes, specifically by
downregulating Bace1 mRNA and upregulating Psen1 mRNA. Heightened β-secretase
activity, part of the amyloidogenic pathway, has been linked to higher amounts of
amyloid deposition. No known mutations exist for β-secretase that predispose
individuals to early-onset AD. Instead, mutations in APP can confer protection or
vulnerability to β-secretase activity. A single residue mutation in APP reduces β-
secretase activity, which has been linked to a reduction in AD risk and cognitive decline
(Jonsson et al., 2012). Several familial APP mutations have amino acid substitutions
that favor catalytic activity by β-secretase, such as the Swedish APP mutation present
in the 3xTg-AD mouse model. High levels of β-secretase have also been shown to be
elevated in late-onset AD, which may be responsible for Aβ deposition in AD
pathogenesis. Therefore, lowered β-secretase activity may be beneficial as it may lead
to less APP processing towards Aβ.
β-secretase cleaves APP to a membrane-bound C-terminal fragment β (CTFβ/C99)
fragment and a soluble extracellular sAPPβ fragment. Subsequent cleavage of CTFβ by
γ-secretase releases the extracellular Aβ monomer. The upregulation of Psen1 with
candesartan, a catalytic subunit of γ-secretase, may be problematic, as it may further
induce generation of Aβ species. However, γ-secretase-mediated Aβ generation is
predicated upon the C99 precursor formed by β-secretase. Therefore, amyloidogenic
128
processing depends on the sequential cleavage of APP through β- and γ-secretases to
release Aβ.
Alternatively, APP can be processed through the non-amyloidogenic pathway via α-
and γ-secretases, which does not yield pathogenic Aβ. Reduced expression of Bace1
mRNA suggests that there may be lowered β-secretase activity associated with
candesartan treatment, and perhaps a shift towards the non-amyloidogenic pathway
through α-secretase. The putative α-secretase gene has not been discovered yet, thus
mRNA expression studies of α-secretase were not feasible. It may be possible to
decipher which pathway is preferred, α- or β-secretase, by examining the soluble
downstream metabolites sAPPα and/or sAPPβ via protein assays.
The downregulation of Bace1 mRNA may be linked to candesartan’s role as a partial
PPAR-γ agonist. It has been shown that PPAR-γ activation represses Bace1 promoter
activity and downstream Bace1 mRNA expression (Sastre et al., 2006). The results
show that only candesartan treatments downregulate cortical Bace1 mRNA, which may
reflect the ability candesartan to effect central PPAR-γ activation.

129

Figure 26 APP Processing Pathways.
The non-amyloidogenic pathway (left) involves cleavage of APP by α- and γ-secretases
to generate sAPPα and C-terminal fragments. The amyloidogenic pathway (right)
involves sequential cleavage by β- and γ-secretases to generate sAPPβ, C-terminal
fragments, and Aβ.

Candesartan and Ang-(1–7) did not affect amyloidogenic APP processing
Since candesartan reduced Bace1 mRNA expression, I surmised that candesartan
may be reducing Aβ production by inhibiting the prerequisite β-secretase step in the
amyloidogenic production of Aβ from APP. To ascertain which APP processing pathway
was preferred, I measured the downstream soluble metabolites of both non-
Aβ
APP
Aβ
β
Aβ
γ
sAPPβ sAPPα
α
γ
extracellular space
cytosol
membrane
amyloidogenic non-amyloidogenic
130
amyloidogenic and amyloidogenic pathways in soluble hippocampal homogenates by
immunoassay: sAPPα and sAPPβ. As this 3xTg-AD mouse model incorporates the
Swedish FAD mutation in APP that preferentially prefers β-secretase cleavage, the
Swedish sAPPβ form was measured.
Despite changes in the transcription of APP processing genes, protein levels of
sAPPα/sw sAPPβ and the ratio of sw sAPPβ-to-sAPPα were unaffected by RAS
treatments. Further, the sw sAPPβ-to-sAPPα ratio was 5:1, indicating that more sAPPβ
was formed compared to sAPPα. This suggests preferential processing toward the
amyloidogenic pathway, characteristic of a pathological Aβ model of AD. These results
suggest that candesartan and Ang-(1–7) did not affect amyloidogenic processing, and
therefore Aβ production, in the 3xTg-AD model.
Candesartan increased the mRNA expression of Aβ-degrading enzymes: IDE,
ECE2, and ACE2
In lieu of effects on Aβ production, the Aβ reduction observed in candesartan- and
Ang-(1–7)-treated 3xTg-AD mice may be attributed through increased Aβ clearance.
Increased Aβ clearance may have been potentially mediated through the upregulation
of Aβ-degrading enzymes: IDE and ECE. IDE, the main soluble Aβ-degrading enzyme,
hydrolyzes multiple peptide bonds of Aβ40 and Aβ42 in the extracellular space (Santos
et al., 2011). To a lesser extent, ECE-1 and -2 may also be involved, considering their
expression was also upregulated by combination treatment. As IDE and ECEs can
degrade monomeric Aβ, the reduction of insoluble Aβ species may be attributed to the
131
increased clearance of soluble species by these monomeric Aβ-degrading enzymes
prior to Aβ fibril and plaque formation.
While ACE expression was thought to be modulated as a major enzyme involved in
both angiotensin and monomeric Aβ catabolism, ACE expression did not appear to be
affected by treatments. Further, the expression of neprilysin, another enzyme
responsible for angiotensin catabolism and plaque degradation, did not appear to be
affected by either Ang-(1–7) or candesartan treatments. Beneficially, both ACE and NEP
were not downregulated with treatments, suggesting preserved expression of these
contributing Aβ-degrading enzymes.
Candesartan treatment also increased ACE2 mRNA expression in 3xTg-AD mice. It
was not surprising, since candesartan has been shown to increase ACE2 expression in
the periphery (Callera et al., 2016). However, the results show that peripheral
administration of candesartan can also increase ACE2 mRNA expression in the brain.
The increased expression of ACE2 in the brain may inhibit the aggregation of certain Aβ
peptides.
The notion that ACE2 may also act as an Aβ-degrading enzyme has only begun to
emerge recently. Recombinant ACE2 was recently described to act upon Aβ43 as a
substrate in vitro, a particularly amyloidogenic Aβ species that seeds plaque formation
(Liu et al., 2014). Conversion of Aβ43 to Aβ42 by ACE2 may allow further proteolytic
degradation by other Aβ-degrading enzymes to lesser neurotoxic forms such as Aβ40.
Since candesartan increased ACE2 mRNA expression in the brain, it is plausible that
candesartan may be inhibiting plaque formation by increasing ACE2 expression to
132
degrade amyloidogenic Aβ43 species. Further research is necessary to determine
whether ACE2 can similarly degrade Aβ43 in vivo, and ACE2’s relative contribution in
the overall pool of Aβ-degrading enzymes. It remains to be seen whether the increased
mRNA expression of these Aβ-degrading enzymes (i.e. IDE, ECEs, and potentially
ACE2) also translates to increased protein expression and functional activity as well.
Candesartan upregulated microglial activation genes
It was postulated that Aβ clearance may also be increasing clearance by activating
glial cells to phagocytose and degrade extracellular Aβ. The results showed that both
candesartan and combination treatments significantly upregulated Iba1 and Cd45
mRNA expression, while Gfap mRNA expression was unaffected. These data suggest
that candesartan activated microglia, but not astrocytes. Microglial activation, or
upregulation of Iba1 and Cd45 mRNA, was likely driven by candesartan, since Ang-(1–
7) by itself did not affect either Iba1 and Cd45. Combination treatment did not appear to
show any synergistic effect beyond a marginal increase in statistical significance.
Unexpectedly, there was no change in cortical Iba1 or Gfap mRNA levels associated
with vehicle-treated 3xTg-AD mice as compared to NonTg controls, despite evidence of
hippocampal microgliosis and astrogliosis in 3xTg-AD mice at as early as 7 months of
age (Caruso et al., 2013). This may be due to less Aβ deposition and neuroinflammation
in the cortex, as compared to the hippocampus in these 13-month-old male 3xTg-AD
mice.
To gauge whether microglia were differentially activated in regards to an M1-M2
phenotype with treatment, M1 and M2 markers along with pro- and anti-inflammatory
133
cytokines associated with activation states were initially investigated in cortical mRNA.
No differences were observed in iNOS, an M1 marker, or pro-inflammatory cytokines IL-
1β, IL-6, or TNF between 3xTg-AD and NonTg mice. Alternate M2 activation markers
exhibited undetectable or inconsistent expression, including M2 markers Arg1, Fizz1,
and Ym1 and anti-inflammatory cytokines Il-13 and Il-10. These data do not suggest the
presence of neuroinflammation or gliosis within the cortex of 13-month-old 3xTg-AD
mice in this study.
Taking these data into consideration, no reasonable inference or conclusion can be
made regarding whether candesartan or Ang-(1–7) treatment polarized microglia to
either activation states or altered cytokine expression profiles. A significant proportion of
these markers were not reliably detectable by qRT-PCR. Another perspective proposes
that M1-M2 phenotypes may not be appropriate in the brain, since much of the research
and literature surrounding M1-M2 activation stemmed from peripheral macrophages and
the assumption that microglia behave similarly (Ransohoff, 2016). However, that is not
to say microglia do not exhibit differential reactive states, since they are known to
surveil their surrounding environment with their processes and adopt amoeboid
morphologies in response to external factors. Contemporary microglia research
suggests that RNAseq and genome-wide expression profiling may be more appropriate
in clarifying microglia activation and function (Ransohoff, 2016).
Whether these activated microglia, induced by treatment, phagocytose Aβ is another
question that gene expression studies cannot answer. Rather, functional phagocytosis
assays or immunofluorescent colocalization studies may be more appropriate and
134
informative to ascertain whether the activated glia truly phagocytosed and degraded Aβ.
These studies are planned.
PPAR-γ Activation
The glial clearance hypothesis stemmed from the prior findings that classical PPAR-
γ agonists such as pioglitazone modulated glial activation and Aβ phagocytosis
(Mandrekar-Colucci et al., 2012). As such, we investigated whether PPAR-γ and LXR
activation genes Abca1, Apoe, Nr1h3, and Pparg were changed with treatment. Both
Ang-(1–7) and candesartan have been previously found to modulate PPAR-γ activation,
but mainly in periphery and not the brain. Within the context of these PPAR-γ/LXR
activation genes, we found that only the combination of candesartan and Ang-(1–7)
increased cortical Pparg mRNA expression compared to vehicle, while other groups
showed no differences. Increased expression of PPAR-γ may, in part, explain the
upregulation of general microglial markers with combination treatment, but does not
account for why candesartan monotherapy also exhibits increased CD45 and Iba1
expression. Since PPAR-γ is a nuclear transcription factor, it may be more appropriate
to investigate the downstream signaling pathways of PPAR-γ activation (e.g. AP2)
rather than PPAR-γ mRNA expression itself. PPAR-γ/LXR target genes LXRα, Apoe,
and Abca1 were not affected, suggesting that these chaperone proteins did not play an
enhanced role in facilitating proteolytic degradation of Aβ.
Candesartan modulates central RAS receptor expression
While RAS-modulating interventions were found to decrease Aβ levels, we sought to
investigate whether changes in central expression of RAS elements also occurred.
135
Cortical brain angiotensinogen mRNA levels were unaffected, but renin levels were
variable, which further supports the notion that renin may not be the functional enzyme
in the brain that cleaves angiotensinogen. Instead, pro-renin and its receptor prorenin
receptor (PRR) may be involved as the main catalytic pathway that drives the
production of angiotensins, but further work must be done to elucidate their role in the
brain (de Kloet et al., 2015).
Candesartan did not alter AT1 expression, which was expected as candesartan acts
as an AT1 antagonist. Further, AT2 expression was also unaltered. However,
candesartan did centrally modulate other elements of the protective RAS, including
Ace2 and Mas1, and unexpectedly Mrgprd and Lnpep.
As described previously, Ace and Ace2 may have a role in degrading monomeric Aβ
species. It was previously shown that candesartan can increase plasma levels of Ace2.
Since candesartan significantly upregulated cortical brain Ace2 mRNA, this suggests
candesartan can penetrate the BBB to affect the central RAS. The downstream effects
of ACE2 may be pleiotropic in the context of AD: it degrades Aβ43 to Aβ42, and
catabolizes Ang II to Ang-(1–7).
Further, candesartan also significantly upregulated Mas1 mRNA, the endogenous
receptor for Ang-(1–7). The Mas receptor has been previously associated with LTP
facilitation, learning, and memory. Thus, increased Ace2 and Mas1 expression suggests
candesartan readily crossed the BBB in 3xTg-AD mice and activated the central
protective Ace2/Ang-(1–7)/Mas pathway. The upregulation of ACE2 and Mas expression
may confer neuroprotective actions by counteracting pathological AT1 signaling to exert
136
anti-inflammatory actions, reduce oxidative stress, promote cerebral blood flow, and
facilitate LTP (Xu et al., 2011).
Unexpectedly, other elements of the RAS may also be involved with candesartan
treatment. As part of an exploratory aim in ascertaining the centrally mediated RAS
effects of candesartan treatment, we sought to describe other pathways, namely
alternative receptors for Ang-(1–7)-like peptides and the AT4 receptor.
It has recently been described that a family of Mas-related G-protein coupled
receptors also bind to Ang-(1–7) and its metabolites. Ala
1
-Ang-(1–7), also known as
alamandine, binds to the Mas-related G-protein coupled receptor member D, encoded
by the Mrgprd gene. Research into the MRGPRD activation begins to suggest that it
elicits effects similar to Mas activation, namely vasodilation and anti-fibrosis. There was
no significant difference in Mrgprd mRNA expression between vehicle-treated 3xTg-AD
mice and NonTg controls. Interestingly, candesartan treatment significantly
downregulated Mrgprd mRNA by at least 1000-fold. The same effect was also seen in
combination with Ang-(1–7), but this effect was likely driven by candesartan since Ang-
(1–7) monotherapy showed no change in Mrgprd expression. The direct significance of
the MRGPRD downregulation is yet to be determined. However, the downregulation of
MRGPRD expression may decrease competitive binding for Ang-(1–7), and increase
available Ang-(1–7) to further bind to the Mas receptor. Thus, the combined effects of
AT1 blockade, ACE2 and Mas upregulation, along with a reduction in MRGPRD
expression, may synergistically stimulate the protective ACE2/Ang-(1–7)/Mas axis in the
brain.
137
Studies have shown that AT4 activation elicits cerebroprotective effects, and may be
involved in blood flow, memory, AD, and Parkinson’s disease. AT4’s role in memory
facilitation, at least in regards to IRAP, is an interesting one. A literature review suggests
that AT4’s memory facilitation properties actually results from the inhibitory actions of
Ang IV on IRAP (Albiston et al., 2011). These results suggest that IRAP inhibition may
underlie the effects of memory improvement seen with Ang IV. Despite this, competing
lines of evidence disagree as to the true identity of the AT4 receptor and whether IRAP
inhibition truly facilitates memory (Wright and Harding, 2011).
Regardless of the true identity of the AT4 receptor, we have shown here that
candesartan significantly upregulated Lnpep mRNA in 3xTg-AD mice by over 1000-fold
compared to vehicle-treated controls. This effect may be related to candesartan’s role
as a partial PPAR-γ agonist, which can sensitize insulin signaling. Insulin has been
found to bring IRAP to the cell surface (Keller, 2004). The significance of this
upregulation of candesartan-induced treatment on Lnpep expression is unclear. In
particular, if IRAP inhibition facilitates memory, then Lnpep upregulation may effect
deleteriously on learning and memory. The involvement of Lnpep is further complicated
by the fact that Lnpep is a pleiotropic aminopeptidase that cleaves multiple substrates,
including: vasopressin, oxytocin, bradykinin, met-enkephalin, dynorphin A, and other
peptide hormones.
It is unclear whether there was any difference in Lnpep expression between 3xTg-
AD mice and their NonTg controls, due to the lack of statistical significance stemming
from the high variance of Lnpep expression in NonTg controls. NonTg controls
138
expressed highly variable Lnpep expression, ranging from 0.2 to 3000-fold change
compared to vehicle-treated 3xTg-AD mice. Further investigation may be warranted into
whether AT4 expression, in terms of Lnpep and the HGF/c-Met receptor system, differ in
AD. It is uncertain whether the upregulation of Lnpep with candesartan treatment is
significant in the context of AD, and whether it prevents or exacerbates cognitive
decline.
In summary, candesartan has been shown to modulate gene expression of the RAS
within the anterior cortex of the mouse brain. In particular, candesartan upregulates
protective elements of the RAS, including Ace2 and Mas. Protective RAS activation may
promote Aβ degradation and learning/memory by facilitating LTP. Candesartan was also
shown to modulate expression of different RAS elements, including the MRGPRD
receptor and IRAP. Whether these findings relate to Aβ reduction is unknown, as this
effect may be only associated with RAS modulation rather than imply any causation in
the context of AD.
139
Summary
It is plausible that the Aβ reduction in 3xTg-AD mice from RAS treatment resulted
from increasing Aβ clearance through Aβ-degrading enzymes. RAS treatments did not
affect amyloidogenic APP processing, as hippocampal protein levels of sAPPα/sAPPβ
remain unchanged. Candesartan and Ang-(1–7) upregulated the mRNA expression of
Aβ-degrading enzymes IDE, ECE2, and ACE2. Microglial activation may also be
involved, as candesartan was shown to increase IBA1 and CD45 expression.
Further work must be conducted to confirm these findings, as mRNA gene
expression does not necessarily correlate to protein expression and function. Further,
this study assessed gene expression only within the cortex, while Aβ reductions were
observed in the hippocampus. Certain genes are differentially expressed within the
brain, such as Mas that is known to be highly expressed within the hippocampus.
Whether these cortical gene expression changes translate to other regions, including
the hippocampus, is unknown. Despite these limitations, these gene expression studies
provide valuable data that may be informative in investigating the mechanisms
underlying Aβ reduction and central RAS modulation from candesartan treatment in
3xTg-AD mice.
140
Chapter 4: EPC Method Development and RAS Receptor Expression
Introduction
Pro-angiogenic hematopoietic cells have been implicated in a number of disease
states, including coronary artery disease (Powell, 2004; Thum et al., 2005; Werner et
al., 2005), rheumatoid arthritis (Grisar et al., 2005), peripheral arterial disease (Shaffer
et al., 2006), diabetes (Segal et al., 2006), chronic obstructive pulmonary disease
(Palange et al., 2006), and Alzheimer’s disease (Kong et al., 2011; Lee et al., 2009;
2010). Many of these pro-angiogenic hematopoietic cells subsets were previously
referred to as endothelial progenitor cells (EPCs), but a recent movement proposes to
clarify these cells based on their methods of isolation, antigen expression, and
proliferative potential (Basile and Yoder, 2014; Fadini et al., 2012; Richardson and
Yoder, 2011).
Characterizing circulating angiogenic cells (CACs) have been difficult due to the lack
of cell surface antigens that can distinguish these cells from other subsets of
hematopoietic cells (Masuda and Asahara, 2003; Rafii and Lyden, 2003). Some studies
have characterized the antigen expression of colony-forming assays and their progeny
(Ahrens et al., 2011; Basile and Yoder, 2014; Desai et al., 2009; Medina et al., 2010;
Yoder et al., 2007), but few have investigated the original cell populations that give rise
to these colonies. Correlating surrogate measures for colony-forming assays would
address some of the limitations of culture-based investigations.
Methods of isolating putative EPCs have been complicated by the heterogeneity of
methods used to measure or enrich EPCs. For example, Lee et al. assessed CACs,
141
CD34+VEGFR-2+, and CD133+ cells in AD patients. In another paper, Kong et al.
evaluated CD34+CD133+ cells in AD patients. Each paper found a reduction in CACs
and CD34+CD133+ cells in AD, respectively, but the results are hard to compare since
different markers were evaluated. I aimed to clarify these populations further by
matching patient samples and characterizing them in parallel by the colony-forming
CFU-Hill assay as well as flow cytometry for all three putative EPC subsets. I propose
that these putative EPC subsets CD34+CD133+ (early progenitors),
CD34+CD133+CD309+ (early endothelial progenitors), and CD34+CD309+ (immature
endothelial progenitors) cell subsets represent sequential stages of progenitor
maturation with differing angiogenic potential.
Modulation of the renin-angiotensin system (RAS) can mobilize these progenitors
and angiogenic cells into circulation. ARBs have been shown to bolster circulating EPCs
in a number of disease states in both humans and animals (Bahlmann et al., 2005;
Endtmann et al., 2011; Pelliccia et al., 2010; Porto et al., 2009; Reinhard et al., 2010;
Townamchai et al., 2010; Wang et al., 2006a; Yao et al., 2007; Yoshida et al., 2011;
You et al., 2008; Yu et al., 2008). Acute AT1 stimulation can initially lead to pro-
angiogenic effects and EPC recruitment, but chronic AT1 stimulation may lead to
diminished EPC numbers and cellular senescence over time (Durik et al., 2012). These
cell subsets may differ in proliferative potential or maturation state based on their RAS
receptor expression. While the RAS is known to modulate EPC numbers, RAS receptor
expression on these cell subsets is unknown.
142
To clarify the role of the RAS in these cells, we evaluated the RAS receptor
expression on these angiogenic cell subsets. Overall, we sought to develop a
reproducible clinical assay to investigate numbers of pro-angiogenic hematopoietic cell
subsets, along with functional assays and RAS receptor expression.
143
Methods
Peripheral Blood Assays
Whole blood from participants (30–50 mL) was collected into K
3
EDTA tubes and
stored on ice for 3–6 hours before processing. Tubes were inverted occasionally every
30 minutes to ensure that parts did not settle. Blood was processed for flow cytometry
(15 mL) and the remainder for CFU-Hill cultures. Total white blood cells were counted
on a hemacytometer using 3% acetic acid with methylene blue.
Blood for flow cytometry and culture assays was diluted 1:1 with Dulbecco’s
phosphate-buffered saline (DPBS) + 2% fetal bovine serum (FBS). Peripheral blood
mononuclear cells (PBMCs) were isolated by density gradient centrifugation with
Histopaque 1077 (Sigma-Aldrich) or Lymphoprep (Stem Cell Technologies) in SepMate
tubes (Stem Cell Technologies). PBMCs were washed twice with DPBS + 2% FBS at
300 × g for 8 min and 120 × g for 10 min (platelet removal) at room temperature (RT).
Blood Smears: WBC Differentials
Blood smears were prepared from a drop of whole blood. Slides were air-dried,
fixed, and stained with Diff-Quik staining kit (Siemens Cat #B4132-1A). Slides were
fixed for 2–3 minutes, then stained and counterstained for 30 seconds each. Random
fields were scored for polymorphonuclear (PMNs) cells and mononuclear cells by a
blinded observer.
Flow Cytometry
PBMCs (1.0 million cells) were transferred to an unstained tube, and the remainder
transferred to a stained tube. The stained tube was incubated with 5 µL of Human BD
144
Fc Block (BD Biosciences) for 10–15 min at RT, and 1 µL of each of the following
antibodies were added: (1) CD34-PE-Vio770 (clone: AC136, Miltenyi Biotec), (2)
CD133-VioBright FITC (clone: AC133, Miltenyi Biotec), (3) CD309-PerCP/Cy5.5 (clone:
7D4-6, BioLegend), (4) angiotensin II type 1a receptor-PE (rabbit polyclonal, Bioss), (5)
AGTR-2-APC (clone: 364805, R&D Systems), and (6) Mas1-Alexa Fluor 350 (rabbit
polyclonal, Bioss). Tubes were incubated in the dark for 30 min at 4 °C. Samples were
washed twice with 3 mL PBS + 2% FBS, centrifuged at 300 × g for 8 min, and fixed with
2% formaldehyde in PBS until analysis.
Samples were acquired on a BD LSR II flow cytometer and analyzed on FlowJo
software. Within the lymphocyte gate, 100,000 events were recorded for the blank tube
and the entire sample was recorded for the stained tube. Fluorescence compensation
was automatically calculated using AbC Total Antibody Compensation Bead Kit (Thermo
Fisher) that captures both mouse and rabbit antibodies. Fluorescence-minus-one (FMO)
controls were used to set positive/negative gates for each fluorochrome in the
lymphocyte gate.
Gating Strategy
Cells were gated to exclude platelets and debris (Region 1 [R1]). A pulse geometry
gate (R2) on FSC-Area vs. FSC-Height was applied to isolate single cells. Within the
lymphocyte gate (R3), single gates were drawn for CD34+ (R4), CD133+ (R5), CD309+
(R6), AT1+ (R7), AT2 (R8), and Mas (R9) above their respective FMO bounds. These
gates were sequentially applied into the lymphocyte gate to assess different cell
subsets, including: CD34+, CD34+CD133+, CD34+CD309+, CD34+CD133+CD309+,
145
as well as their RAS receptor expression. Percentage of RAS receptor expression on
various cell subsets were only quantified in our statistics if that cell population had at
least 10 cells. For example, CD34+CD133+CD309+ cells are typically very rare; 2 triple-
positive cells would not be assessed for RAS receptor expression.
Progenitor Cell Enrichment for Fluorescence-Minus-One Bounds
To determine the fluorescence-minus-one (FMO) boundaries for gating EPCs,
PBMCs were enriched for progenitor cells. Human progenitor cells were obtained from
80 mL of donor peripheral blood using an immunomagnetic negative selection kit from
Stem Cell Technologies (Catalog #19056).
Enriched progenitor cells were split into FMO tubes, and each tube was stained
using every fluorochrome-conjugated antibody in the EPC flow cytometry protocol
except the fluorochrome of interest. Refer to the Table of FMO controls.
Within the lymphocyte gate, FMO bounds were established as a reference gate
compared to the unstained tube for each antigen of interest. These representative FMO
bounds over unstained bounds were used to establish the flow cytometry gates for the
rest of the study.
The overall gating strategy was to gate for an individual channel of interest against
another channel with minimal spillover. For example, CD34-PE-Vio770 was gated
against the CD133-FITC channel and vice versa. CD309-PerCP/Cy5.5 was gated
against CD133-FITC channel.

146
Antigen
/Fluorophore
VioBright
FITC
PE PerCP/
Cy5.5
PE-
Vio770
APC Alexa
Fluor 350
Unstained ---- ---- ---- ---- ---- ----
CD133 FMO ---- AT1 CD309 CD34 AT2 Mas
AT1 FMO CD133 ---- CD309 CD34 AT2 Mas
CD309 FMO CD133 AT1 ---- CD34 AT2 Mas
CD34 FMO CD133 AT1 CD309 ---- AT2 Mas
AT2 FMO CD133 AT1 CD309 CD34 ---- Mas
Mas FMO CD133 AT1 CD309 CD34 AT2 ----

Table 5 Panel of fluorescence-minus-one (FMO) controls.
FMO controls were used to establish gating boundaries in enriched human progenitor
cells, stained for every fluorochrome-conjugated antibody except for the fluorochrome of
interest.

Colony-Forming Assays
CFU-Hill Assay
PBMCs were processed for the CFU-Hill assay using a commercially available kit
(Stem Cell Technologies). Briefly, PBMCs were seeded on 12-well fibronectin-coated
plates (Corning) at 2.5 × 10
6
cells/well in CFU-Hill Liquid Medium (formerly known as
EndoCult) and incubated for 2 days at 37 °C. Non-adherent cells were collected and
replated onto 24-well fibronectin-coated plates at 1.0 × 10
6
cells/well and incubated for 3
days at 37 °C. At day 5, CFU-Hill colonies in each well were scored as a central core of
round cells surrounded by radiating spindle-shaped cells (Hill et al., 2003). Functional
147
assays were conducted on these wells at day 5 including: senescence-associated β-
galactosidase staining, AcLDL uptake, and conditioned media.
Circulating Angiogenic Cells (CFU-EPCs)
Cultures of circulating angiogenic cells (CACs), also known as CFU-EPC, were
processed according to a previously published method (Lee et al., 2009). Briefly,
PBMCs were resuspended in endothelial growth medium-2 BulletKit (CC-3162, Lonza)
and seeded onto 12-well fibronectin- and 2%-gelatin-coated plates at 2.5 × 10
6

cells/well. Coated plates were prepared to achieve a surface area coverage of 1 µg/cm
2

fibronectin or 0.2 mg/cm
2
gelatin (Sigma Cat #F0895, G1393). Cells were incubated in a
humidified 37 °C chamber with 5% CO2 for 7 days. At day 7, CFU-EPC colonies in each
well were scored as a central core of round cells surrounded by radiating spindle-
shaped cells (Hill et al., 2003).
Functional Assays
Senescence-associated β-galactosidase staining
Senescence-associated β-galactosidase (β-gal) staining in CFU-Hill cells were
assessed using a Cellular Senescence Assay Kit (Marker Gene Technologies). At day 5
of the CFU-Hill assay, one well was fixed, washed, and incubated with 1 mg/mL X-Gal
for 16–20 hours. Cells were examined and photographed using an inverted microscope.
Only isolated spindle-shaped cells distant from central colonies were analyzed, and
cells with a distinctly blue cytoplasm (indicating β-gal activity) were counted. The % β-
gal+ cells was determined by counting at least 3 random fields (Assmus et al., 2003; Hill
et al., 2003; Lee et al., 2010).
148
AcLDL Uptake
At day 5 of the CFU-Hill assay, the media from one well was replaced with Medium
199 containing 10 µg/mL Alexa Fluor 488-conjugated AcLDL and incubated for 3–4
hours at 37 °C (Shaffer et al., 2006). Cells were washed, trypsinized, and fixed in 2%
formaldehyde. Samples were acquired on a BD LSR II flow cytometer and analyzed for
% AcLDL+ cells on FlowJo software.
Boyden chamber chemotaxis assay
The chemotactic ability of CFU-Hill cells at day 5 was assessed using the CytoSelect
Boyden chamber chemotaxis assay (Cell Biolabs). Briefly, cultured CFU-Hill cells at day
5 were detached using 1 mM EDTA in PBS, harvested by centrifugation, and
resuspended in Medium 199. 1.0 x 10
5
cells were placed in the upper 8 µm-pore insert,
and then placed into the lower well of a 24-well plate containing 50 ng/mL VEGF. After
incubating at 37 °C for 24 hours, media were removed from the insert. Non-migratory
cells were removed by swabbing the interior of the inserts. Remaining cells will be fixed
and stained. Migratory cells were counted in three random fields using an inverted
microscope (Lee et al., 2010).
Matrigel Tube Formation Assay
Angiogenic activity of conditioned media samples were assessed using the In Vitro
Angiogenesis Assay Kit for Tube Formation (Cultrex), adapted from a previous method
(Lee et al., 2010). Conditioned media for this assay were collected by growing CFU-Hill
cells for 5 days, replacing the media with serum-free Medium 199, and then collecting
the conditioned media overnight. A 96-well tissue culture plate was coated with Matrigel
149
and incubated at 37 °C overnight. EAhy.926 cells were grown overnight in DMEM +
10% FBS, trypsinized, and resuspended in Medium 199, CFU-Hill Liquid Medium, or
conditioned media at 1.5 x 10
5
cells/mL. 15,000 cells/100 µL was added to each well on
the Matrigel-coated plate. After 24 hours, the medium was removed and the cells were
fixed, stained, and visualized using an inverted microscope.
Statistical Analyses
Statistical analyses were performed using GraphPad Prism 6 and SPSS 24.0. Tests
were considered significant if they met a significance threshold (p value) below 0.05.
Continuous variables were assessed for normality using skewness, kurtosis, and the
Shapiro-Wilk test. Parameters were log10-transformed as appropriate to satisfy the
assumption of normality for parametric tests.
Participant Characteristics
Participant characteristics were assessed based on demographics, medical history,
and current prescriptions. Framingham risk scores for stroke and cardiovascular
disease (office-based BMI method) were calculated accordingly (D'Agostino et al., 1994;
2008). Framingham risk scores correspond to a 10-year stroke or cardiovascular
disease event probability based on data from the Framingham heart study. The systolic
blood pressure was taken from the mean of arm cuff measurements taken during two
clinical study visits.

150
Factors Influencing Angiogenic Cell Subsets
Correlations to CFU-Hill
We evaluated whether CFU-Hill colonies correlated with flow cytometry (FC) subsets
and hematological parameters. Previous studies have shown that CFU-Hill colonies
correlate with Framingham risk score, total leukocytes, and monocytes (Hill et al., 2003;
Prokopi et al., 2009). As these confounding factors are known to influence angiogenic
cell subsets, we sought to evaluate whether these factors significantly altered
angiogenic cell numbers in this dataset.
Pearson correlation coefficients (r) were obtained for CFU-Hill colonies versus FC
subsets, total WBC counts, % polymorphonuclear cells, and % mononuclear cells. The
Spearman correlation coefficient (ρ) was obtained for CFU-Hill colonies and
Framingham risk scores (Lee et al., 2009).
RAS receptor expression on angiogenic cell subsets
The RAS receptor expression was enumerated on varying subsets of the triple
putative EPC markers CD34, CD133, and CD309. These subsets included: CD34+,
CD133+, CD309+, CD34+CD133+, CD34+CD133+CD309+, and CD34+CD309+ cells.
Percentage of RAS receptor expression on cell subsets were only quantified in the
statistics if that cell population had at least 10 cells. Medians and interquartile ranges
were obtained for RAS receptor expression.
A non-parametric Friedman test with Dunn’s multiple comparisons test was
conducted to assess whether there were any significant changes in RAS receptor
expression between matched CD34+CD133+, CD34+CD133+CD309+, and
151
CD34+CD309+ subsets. Only samples with enumerable counts in all three cell subsets
were included to avoid bias.
152
Results
EPC Method Development
A polychromatic flow cytometry protocol was developed to detect pro-angiogenic
hematopoietic cell subsets and their RAS receptor expression. The antibody panel
included: CD34, CD133, CD309, AT1, AT2, and Mas. Some of these cell subsets (e.g.
CD34+CD133+, CD34+CD133+CD309+, CD34+CD309+) have previously been
referred to as putative markers for EPCs. Running compensation and FMO controls was
necessary to minimize the effects of fluorescence spillover.
All angiogenic cell subsets were detected in all 31 blood samples successfully
processed for flow cytometry. CFU-Hill colonies were successfully obtained for 27
samples. Neither angiogenic cell subsets processed for flow cytometry or colony-
forming assays satisfied the assumption of normality.
Despite previously published reports, putative EPC colonies could not be grown
consistently according to the CAC assay (Lee et al., 2009). In a set of pilot experiments,
the CAC assay showed little-to-no colony formation when grown in EGM-2 media for 7
days on either fibronectin- or gelatin-coated plates. Varying conditions were utilized,
including: no media changes, media changes every 1–3 days, and replating non-
adherent cells after 1–2 days.
To achieve a feasible and reproducible EPC colony-forming assay, we explored
using a commercially available kit, the CFU-Hill Liquid Medium Kit from Stem Cell
Technologies. Using the CFU-Hill assay, we successfully cultured CFU-Hill colonies
from PBMCs obtained from both younger controls and older participants in 5 days.
153
Thus, the colony-forming assay portion of this study was conducted using the
commercial CFU-Hill assay instead of the originally proposed CAC assay.
As proof of concept, we seeded PBMCs in CFU-Hill Liquid Medium and EGM-2
using a matched sample from an older adult. The CFU-Hill assay showed ample colony
growth at day 5 (mean CFU = 16.8, N = 6 wells). Cells grown in EGM-2 on gelatin-
coated plates showed no colony growth after 7 days (mean CFU = 0, N = 21 wells).
Cells in EGM-2 were grown out further for an additional 12 days, with media changes
every 3 days. At day 19, no colonies were observed in cells grown in EGM-2 on gelatin-
coated plates.
Progenitor Enrichment & FMO Bounds
Enriched PBMCs yielded 6 times higher CD34+ cells versus normally isolated
PBMCs by density gradient centrifugation (0.43% CD34+ enriched cells vs. 0.07%
unenriched PBMCs). Both enriched progenitor cells and normal PBMCs were stained
for FMO controls. Positive gates for each antigen were established relative to the clearly
delineated FMO bounds from the enriched progenitor cells.
FMO bounds were established as a reference gate compared to the unstained tube
for each antigen of interest. These representative FMO bounds over unstained bounds
were used to establish the flow cytometry gates for the rest of the study.

154
Unenriched PBMCs
15 mL Blood
HPC Enrichment
80 mL Blood
Morphology
FSC vs. SSC

CD34/CD133

CD309

Figure 27 FMO boundaries from enriched progenitor cells.

155
Colony-Forming Assays
(a)
(b)
Figure 28 Morphology of Colony-Forming Assays.
Representative micrographs from (a) CFU-Hill assay at day 5 and (b) CAC assay at day
7. The CFU-Hill assay formed discrete CFU-Hill colonies, consisting of a central core of
rounded cells with radiating spindle-shaped cells in the periphery. The CAC assay failed
to form colonies.
Functional Assays
β-Gal
SA-β-gal staining was conducted on a representative well from the CFU-Hill assay,
and reliably detected in 20 of 27 CFU-Hill assays. The SA-β-gal assay was omitted if
156
there was no colony formation or insignificant numbers of spindle-shaped cells to score.
% β-gal+ cells were normally distributed according to the Shapiro-Wilk test.

Figure 29 SA-β-Gal Staining of CFU-Hill Assay.
Photographs of spindle-shaped cells distant from central colonies were captured for β-
gal scoring.

AcLDL uptake requires 3–4 hours incubation
To ensure consistency from sample to sample, as samples were processed
sequentially, I conducted a validation study to assess AcLDL uptake at various time
points. Some participant samples were incubated for less than 4 hours for AcLDL
uptake. A time validation assay was conducted to validate the appropriate minimum time
of incubation for significant AcLDL uptake.
CFU-Hill colonies were obtained by growing PBMCs in CFU-Hill Liquid Medium for 5
days at 37 °C. One participant showed an abundance of CFU-Hill colonies with plentiful
spindle-shaped cells and low variance. Mean colony number ± SD was 21.9 ± 4.0 CFUs
per million cells.
157
In each well, adherent cells were washed and incubated with 10 µg/mL Alexa Fluor
488-AcLDL in Medium 199 (basal medium) for 1, 2, 3, and 4 hr at 37 °C in duplicate.
Cells were trypsinized, washed, and fixed for flow cytometric analysis. Stained AcLDL+
cells were manually gated above the autofluorescence of blank cells in the FITC
channel.
After 1 hour of incubation, 30–40% of cells ingested AcLDL with 1 hour of incubation,
and the uptake increased with time up to 90% after 3–4 hours of incubation. Since
colony number variance was low, I concluded that cells must be incubated with 10
µg/mL AcLDL at 37 °C for 3–4 hours for consistent, longitudinal results.
It should also be noted that fluorophore-conjugated AcLDL degrades with time, and
the manufacturer suggests a shelf life of 2–3 months (Cat #L23380, Thermo Fisher).

Figure 30 Flow Cytometry of AcLDL Uptake in Putative EPCs.
158

Figure 31 AcLDL uptake required 3–4 hours of incubation.
CFU-Hill colonies readily take up AcLDL
CFU-Hill colonies readily take up AcLDL. Cultures that exhibited the characteristic
CFU-Hill morphology (i.e. rounded cells surrounded by radiating spindle-shaped cells)
had a median of 86% AcLDL+ cells. CFU-Hill colony number did not correlate with
AcLDL uptake (r = 0.023, p = 0.929, N = 18).

Figure 32 CFU-Hill colony number did not correlate with AcLDL uptake.
159
Correlations to CFU-Hill
CD34+CD133+ cells correlated with CFU-Hill colonies
CFU-Hill colony number significantly correlated to three FC subsets: CD34+CD133+
(p < 0.0001), CD34+CD133+CD309+ (p < 0.05), and CD34+CD133+CD309- (p <
0.0001), as determined by Pearson correlation. Correlations with other FC subsets were
not found to be statistically significant.

Cell Subsets r p value
CD34+ 0.318 0.09940
CD133+ 0.109 0.58121
CD309+ -0.002 0.99105
CD34+CD133+ 0.698 0.00004
CD34+CD133+CD309+ 0.405 0.03274
CD34+CD309+ 0.096 0.62674
CD34+CD133+CD309- 0.702 0.00003
CD34+CD133-CD309+ 0.059 0.76553
% EPCs in CD34+ 0.153 0.43764

Table 6 Correlations between CFU-Hill colonies and flow cytometric cell subsets.
Pearson correlation coefficients (r) were obtained for CFU-Hill colonies versus flow
cytometric subsets. Cell subsets were normalized to 1 million cells.
160

Figure 33 Circulating CD34+CD133+ cells significantly correlated with CFU-Hill
colony number.
Pearson correlation coefficient r = 0.698, n = 28, p < 0.0001. CFU-Hill colonies were
expressed as the mean number of colonies per 1 million cells plated per well, and
CD34+CD133+ cells were normalized per 1 million cells.

CFU-Hill colonies did not correlate with Framingham risk scores
No significant correlations were observed between CFU-Hill colonies and
Framingham stroke and CVD risk scores. There was a negative trend between CFU-Hill
colonies and Framingham CVD risk score, although it was statistically insignificant (ρ = -
0.350, p = 0.079).

161
Framingham risk score ρ p value
Stroke -0.251 0.216
CVD -0.350 0.079

Table 7 Correlation between CFU-Hill colonies and Framingham risk scores.
Spearman’s rank correlation coefficients (ρ) were obtained for CFU-Hill colonies versus
Framingham stroke and cardiovascular (CVD) risk scores.

CFU-Hill colonies did not correlate with hematological parameters
CFU-Hill colonies did not correlate significantly to total WBCs, % PMNs, or % MNCs.
Hematological parameters were normally distributed with no outliers outside of normal
physiological ranges.

r p value
Total WBCs -0.11 0.59
% PMNs 0.26 0.22
% MNCs -0.26 0.22

Table 8 Correlation between CFU-Hill colonies and hematological parameters.
Pearson correlation coefficients (r) were obtained for CFU-Hill colonies versus total
white blood cells (WBCs), % polymorphonuclear cells (PMNs), and % mononuclear
cells (MNCs).

162
CD34+CD133+CD309+ cells express CD34
bright
population
Within the CD34+CD309+ gate, the CD133 gate delineates the boundary between
CD34+CD309+CD133+ and CD34+CD309+CD133- cells. A bivariate plot (CD34 vs.
CD133) within the CD34+CD309+ population revealed a distinct CD34
bright
population
within CD133+ cells, shown in the following figure.

Figure 34 CD34
bright
population in CD34+CD133+CD309+ subsets.
A representative bivariate plot within CD34+CD309+ cells showed a distinct CD34
bright

population in CD133+ cells. This pseudocolor plot depicts high cell density as red-
orange areas.

163
RAS receptor expression on angiogenic cell subsets
RAS receptor expression was compared between putative EPC subsets
CD34+CD133+ (early progenitors), CD34+CD133+CD309+ (early endothelial
progenitors), and CD34+CD309+ (immature endothelial progenitors) cells. Receptor
expression on cell subsets were compared to CD34+CD133+CD309+ cells as the
comparator group. 97% of CD34+CD133+CD309+ cells were AT1+, which was
significantly higher than CD34+CD133+ (59% AT1+) cells and CD34+CD309+ (52%
AT1+) cells. AT2+ cells increased from CD34+CD133+ (1%) to CD34+CD133+CD309+
(9%) to CD34+CD309+ (22%). Mas expression was low and did not significantly change
across subsets (0–1.5%).

Figure 35 RAS receptor expression on cell subsets.
Data are shown as median with interquartile ranges.
Mononuclear Cells
CD34+
CD133+
CD309+
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
0
20
40
60
80
100
% Receptor Expression
AT1
Mononuclear Cells
CD34+
CD133+
CD309+
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
1
10
100
% Receptor Expression
AT2
Mononuclear Cells
CD34+
CD133+
CD309+
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
1
10
100
% Receptor Expression
Mas
164

Figure 36 RAS receptor expression on putative EPC subsets.
The expression of AT1, AT2, and Mas receptors was characterized on putative EPC
subsets: CD34+CD133+, CD34+CD133+CD309+, and CD34+CD309+. Data are shown
as median with interquartile ranges. * p < 0.05, ** p < 0.01, **** p < 0.0001.

CD34+
CD133+
CD34+
CD133+
CD309+
CD34+
CD309+
% AT1 59.42 **** 97.62 51.56 ****
% AT2 1.16 ** 8.7 22.39 *
% Mas 0.18 0 1.54

Table 9 RAS receptor expression on putative EPC subsets.
The expression of AT1, AT2, and Mas receptors was quantified on CD34+CD133+,
CD34+CD133+CD309+, and CD34+CD309+ cells. Statistical significance of receptor
expression was compared to CD34+CD133+CD309+ cells. * p < 0.05, ** p < 0.01, **** p
< 0.0001.
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
0
20
40
60
80
100
AT1
% Receptor Expression
****
****
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
0
10
20
30
40
50
AT2
% Receptor Expression
**
*
CD34+ CD133+
CD34+ CD133+ CD309+
CD34+ CD309+
1
10
Mas
% Receptor Expression
165
Discussion
Pro-angiogenic hematopoietic cell subsets have been implicated, associated, and
dysregulated in multiple disease states and may serve as an indicator of cardiovascular
risk. Different methods of isolating, enumerating, and proliferating these angiogenic cell
subsets complicate research findings, leading to uncertainty as to whether results from
different methods yield comparable and valid findings. Two commonly used techniques
in measuring angiogenic cell subsets include: flow cytometry of putative EPC cell
subsets and colony-forming assays. To see whether these cell subsets relate to each
other, I developed a robust and reproducible clinical protocol to investigate angiogenic
cell subsets by flow cytometry and CFU-Hill colonies in parallel.
Assay Development Considerations
The characterization of angiogenic cell subsets through flow cytometry and colony-
forming assays confers different types of information. Flow cytometry assesses cellular
antigen expression, whereas colony-forming assays can be informative for proliferative
potential and function. In other words, flow cytometry can assess markers associated
with angiogenesis, whereas cultures may be more informative in terms of angiogenic
potential through colony formation. Cultured cells and conditioned media can be
collected for functional assays, including gene expression, cytokine release, in vitro tube
formation, etc.
It is prudent to choose bright fluorophores (FITC, PE, PE-Cy7, APC) with distant
excitation/emission spectra for examining rare cells, such as pro-angiogenic
hematopoietic cells or putative EPCs (Pober, 2012). While this does not negate the
166
necessity of proper compensation and FMO controls, this limits fluorescence spillover.
Future studies may also consider using a dump channel to exclude undesired cells
through negative selection, such as dead cells using a viability dye or mature leukocyte
markers such as CD45. A dump channel can encompass multiple negative selection
markers using the same fluorophore and filter set. Dead cells can be particularly
problematic, as they can nonspecifically bind to antibodies and give rise to inaccurate
results.
The feasibility of colony-forming assays is a concern, particularly in terms of length
of culture and the possibility of contamination. Antibiotics can mitigate contamination to
a degree, but it has been shown that antibiotics can suppress hematopoietic colony
formation (Maruyama et al., 1987). Therefore, I have elected to not use antibiotics in our
assays, but this has resulted in the absence of results in some samples due to microbial
contamination. In contrast, flow cytometry is quicker and less prone to contamination.
Characterizing angiogenic cells through flow cytometry and colony-forming assay in
parallel mitigates the shortcomings of either assay alone.
Functional Assays
Functional assays were conducted for culture studies, per a prior study
demonstrating dysfunctional characteristics of angiogenic cells in AD patients (Lee et
al., 2010). Cells from AD patients expressed increased senescence and impaired
angiogenic activity on the Boyden chemotaxis and Matrigel tube formation assays.
Senescence-associated β-galactosidase (SA-β-gal) staining of cells was evaluated
in the CFU-Hill assay. Spindle-shaped cells stained distinctly blue were scored as SA-β-
167
Gal+. This assay was validated for 16–20 hours of incubation, as further incubation may
elicit false positives. CFU-Hill assays that yielded insignificant numbers of colonies or
spindle-shaped cells were not able to be scored. Thus, the SA-β-gal assay depends on
successful colony formation and an ample number of spindle-shaped cells.
Initially we conducted the AcLDL uptake assay to confirm endothelial phenotype of
CFU-Hill colonies. Previous methods have incubated AcLDL for 1–4 hours (Lee et al.,
2010; Shaffer et al., 2006). To ensure consistency for the longitudinal nature of this
study, a validation study was conducted in a matched sample to investigate whether
incubation times affect AcLDL uptake. Lower incubation times (e.g. 1 hour) showed
incomplete AcLDL uptake. Results showed that 3–4 hours of incubation time were
necessary to characterize AcLDL uptake consistently.
AcLDL uptake results did not show any significant findings in the context of our
parameters of interest (e.g. cognitive status or MRI measures), and did not correlate to
CFU-Hill colony number. Past EPC studies have utilized AcLDL uptake to confirm
endothelial phenotype, but this may not be a specific finding, since monocytes can also
uptake AcLDL. In addition, due to the lengthy 4-hour incubation and the short shelf-life
of AcLDL reagents, this assay was discontinued.
The Boyden chamber chemotaxis assay was conducted on CFU-Hill cells to
characterize chemotactic ability. However, low cell yield from individual wells prevented
further characterization.
I conducted pilot studies to assess whether CFU-Hill cells form tubes on the Matrigel
assay. No tube formation was observed with CFU-Hill cells, in line with prior findings
168
with other groups. CFU-Hill cells do not form tube-like structures on Matrigel when
cultured alone, but may exert an angiogenic stimulatory effect on the ability of
endothelial cells to form tubules (Ahrens et al., 2011). I confirmed that the EA.hy926 cell
line readily forms tubes when grown on Matrigel in Dulbecco’s Modified Eagle’s Medium
(DMEM) + 10% FBS (Unger et al., 2002). It may be possible to assess the angiogenic
potential of conditioned media obtained from the CFU-Hill assay along with an
endothelial cell line that readily forms tubes on Matrigel such as EA.hy926 or HUVEC.
However, these conditioned media studies with cell lines should be conducted
simultaneously, since cells can change phenotype based on passage.
CD34+CD133+ cells correlated to CFU-Hill colonies
CD34+CD133+ cells significantly correlated positively with CFU-Hill colonies (r =
0.698, p < 0.0001), which suggests they may derive from a similar hematopoietic
population. This finding increases the validity of comparing results from clinical studies
investigating either CD34+CD133+ cells by flow cytometry or CFU-Hill colonies.
For example, two separate studies found reductions in CD34+CD133+ cells and
colony-forming-unit EPCs in patients with Alzheimer’s disease (Kong et al., 2011; Lee et
al., 2009). Since CD34+CD133+ cells correlate with CFU-Hill colonies, these studies
altogether provide stronger evidence for a deficit in pro-angiogenic hematopoietic cells
in AD. This suggest that CACs/CFU-Hill colonies and CD34+CD133+ cells may arise
from a similar hematopoietic lineage that may be implicated in the pathogenesis of AD.
CFU-Hill colonies also correlated to CD34+CD133+CD309+ (r = 0.405, p < 0.05)
and CD34+CD133+CD309- (r = 0.702, p < 0.0001), but this effect is likely driven by the
169
parent CD34+CD133+ population. No correlation was observed for CD34+CD309+
subsets.
While CD34+CD133+ correlated strongly to CFU-Hill colonies, CD34+CD133+ cells
were not completely predictive of CFU-Hill colony formation. Despite the significant
correlation, outliers were clearly evident. Thus, a linear regression model using
CD34+CD133+ to predict CFU-Hill colony formation may be not be appropriate.
Correlation of CFU-Hill to other parameters
Past studies have shown CFU-Hill colonies to correlate with other parameters such
as Framingham cardiovascular risk score, monocytes, and total WBCs (Hill et al., 2003;
Prokopi et al., 2009). No correlation of CFU-Hill was observed in our dataset versus
Framingham risk scores, total WBCs, PMNs, or MNCs, but this might be due to the low
sample size (n = 27) and normal distribution. Hematological parameters (WBC count,
PMNs, and MNCs) were within normal physiological ranges.
EPC Maturation Stages and RAS Receptor Expression
Based on these data and the literature, CD34+CD133+ and CFU-Hill colonies
represent a population of pro-angiogenic hematopoietic cells. While neither
CD34+CD133+/-CD309+/- cells nor CFU-Hill colonies fit the criteria for a true EPC,
clearly these cell subsets have been shown to be implicated in an array of diseases and
cardiovascular risk.
Expression of CD34 and CD133, hematopoietic progenitor markers, diminish with
maturation and differentiation. CD133 was first discovered to be co-expressed with
CD34
bright
cells from human fetal liver, BM, and blood (Miraglia et al., 1997; Yin et al.,
170
1997). Similarly, I observed a CD34
bright
population within CD34+CD133+CD309+ cells.
This CD34
bright
population saturated the fluorescence channel, indicating high CD34
expression on these cells. While potentially a compensation or spillover artifact, this was
unlikely considering the distinct excitation (Ex) and emission (Em) spectra for CD34-PE-
Vio770 (Ex/Em [nm]: 565/775) and CD133-VioBright FITC (Ex/Em [nm]: 496/522)
showed little overlap. CD34
bright
CD133+ cells in cord blood have been shown to exhibit
high proliferative potential (Goussetis et al., 2000). This suggests that CD34+CD133+
cells represent a population enriched for early progenitors with high proliferative
potential.
I show here that angiogenic cell subsets CD34+CD133+, CD34+CD133+CD309+,
and CD34+CD309+ exhibited high AT1 receptor expression. In particular, AT1
expression was significantly higher in CD34+CD133+CD309+ cells compared to the
other two subsets. The differential expression of AT1 receptor may be related to the
angiogenic potential of these subsets.
Prior studies have shown that Ang II induced proliferation and differentiation of CACs
by upregulating CD309/VEGFR-2 expression (Imanishi et al., 2004). These effects were
blocked by valsartan, which suggests these angiogenic effects were mediated through
Ang II’s action on the AT1 receptor. It is conceivable then that the upregulation of AT1
receptor expression may be associated with increased angiogenic potential on these
cell subsets.
However, increased AT1 expression may be detrimental as well. While AT1
stimulation can initially lead to pro-angiogenic effects and EPC recruitment, chronic AT1
171
activation has been shown to increase oxidative stress and promote EPC senescence
and apoptosis (Durik et al., 2012). It remains to be seen whether chronic AT1
stimulation in these pro-angiogenic hematopoietic populations exhausts proliferative
ability or increases senescence, but further mechanistic studies are warranted.
Interestingly, AT2 expression increased from CD34+CD133+ (1% AT2+), to
CD34+CD133+CD309+ (9% AT2+), to CD34+CD309+ (22% AT2+) subsets. AT2 may
have a role in the differentiation of angiogenic cells, as it has been shown that AT2
blockade with PD123,319 inhibited Ang II-stimulated endothelial cell differentiation of
mesenchymal stem cells (Ikhapoh and Pelham, 2015). Multiple studies reaffirm this
finding wherein AT2 blockade inhibits Ang II-induced endothelial differentiation,
suggesting that that AT2 plays a critical role in differentiation and maturation of
endothelial cells (Walther et al., 2003). Therefore, CD34+CD309+ cells may represent a
more mature, differentiated population based on their AT2 expression.
Mas expression was observed to be low in these angiogenic cells subsets, with no
significant difference in Mas immunoreactivity between subsets. This suggests that Mas
may not be involved in maturation or angiogenic processes of these angiogenic cells.
However, it is possible that Mas expression was not reliably detected. Optimization
assays (data not shown) displayed significantly higher expression in a Mas-transfected
human embryonic kidney HEK293 cell line when permeabilized, compared to not. It is
possible that the Mas antibody in this flow cytometry panel targeted an internal epitope,
thereby precluding reliable detection of Mas expression. Reliable detection of Mas
expression by flow cytometry may require permeabilization, a validated antibody that
172
recognizes an external epitope, or alternatively a fluorescent reporter that targets Mas
mRNA.
Based on these results, I propose that CD34+CD133+, CD34+CD133+CD309+, and
CD34+CD309+ cell subsets represent sequential stages of progenitor maturation with
differing angiogenic potential. CD34+CD133+ cells contain an enriched population of
early progenitors with high proliferative potential. Commitment of progenitors toward
endothelial lineage is associated with increased expression of CD309/VEGFR-2
coinciding with high expression of AT1. As these cells mature, they lose the early
progenitor marker CD133 and downregulate AT1 expression. Along this maturation
pathway, AT2 expression steadily increases.
While these cell subsets have not been shown to differentiate into endothelial cells,
they may have a role as cells supporting angiogenesis. Evidence suggests that
endothelial cell differentiation depend on the opposing actions of the AT1/AT2 axis
(Ikhapoh and Pelham, 2015; Walther et al., 2003), upon which these angiogenic cell
subsets clearly show differential RAS receptor expression.
It is acknowledged that these observations are limited in scope, as it is a cross-
sectional study of cells in peripheral blood. Hematopoietic progenitor cells are rare in
circulation, and require a large volume of blood for enumeration and analysis of cell
subsets. Other sources enriched in progenitors may be utilized to elucidate these cell
subsets, including: mobilized peripheral blood, cord blood, and BM. It should be noted
the markers used for angiogenic cell subsets represent progenitors (CD34, CD133) or
early endothelial markers (CD309/VEGFR-2). While CD309/VEGFR-2 is a marker of
173
endothelial lineage, high expression represents an immature endothelial phenotype.
Mature endothelial markers may be more appropriate to ascertain later maturation
stages, such as von Willebrand factor (vWF), CD146, or vascular endothelial cadherin
(VE-cadherin), in combination with mature endothelial cells such as HUVECs (Goon et
al., 2006; Nguyen et al., 2016). Further work is justified to elucidate the role of the RAS
in these angiogenic cells.
Summary
In summary, I have developed a clinical protocol to characterize angiogenic cell
subsets (CD34/CD133/CD309) and their RAS receptor expression (AT1/AT2/Mas) by
flow cytometry and the colony-forming CFU-Hill assay in peripheral blood.
CD34+CD133+ cells significantly correlated positively to CFU-Hill colonies, suggesting
they may derive from a similar hematopoietic population, which provides further validity
to comparing studies investigating only one or the other. Within the flow cytometric
angiogenic cell subsets, differential RAS receptor expression was observed in regards
to AT1-AT2 expression. These findings imply a possible maturation pathway along with
differing stages of angiogenic potential.
174
Chapter 5: Clinical Study of Angiogenic Cells in MCI
Introduction
Studies have found that circulating pro-angiogenic progenitor cells, previously
referred to as EPCs, were diminished and dysfunctional in AD (Kong et al., 2011; Lee et
al., 2009; 2010). The reduction in these cell subsets, designated by the expression of
CD34, CD133, CD309/VEGFR-2, or colony-forming assays, also correlated to reduced
cognitive performance. These results suggest abnormal vascular repair mechanisms
are associated with AD.
However, it is unknown what role these vascular repair mechanisms play in cognitive
decline prior to AD. Further studies are needed to determine the role of these cell
subsets in normal cognition and MCI, and whether a reduction in circulating pro-
angiogenic progenitor cells could be a prognostic marker for AD.
I hypothesized that prior to AD, cognitive decline coincides with the depletion of
reparative pro-angiogenic progenitor cells. As part of an observational clinical research
study, circulating pro-angiogenic progenitor cell subsets were examined in older adults
with normal cognition and MCI. These cell subsets were correlated to cognitive scores
and volumetric MRI measures.
175
Methods
Participant Recruitment
The protocol for the study was approved by the Institutional Review Board at the
University of Southern California (USC) and written informed consent was obtained from
all subjects. Subjects were recruited through the Vascular Senescence and Cognition
(VaSC) Lab at USC by word of mouth, flyers, and community outreach events.
Participants were 55 years of age or older and living independently. Subjects underwent
a comprehensive neuropsychological assessment, MRI brain scan, and blood draw.
Subjects fasted overnight prior to the blood draw.
Neuropsychological assessment, MRI, and phlebotomy was conducted by the VaSC
Lab. Blood was transported to the USC School of Pharmacy, where the peripheral blood
assays, including plasma isolation, flow cytometry, and cell culture were conducted.
Peripheral blood assays were conducted in a blinded fashion without knowledge of the
participants’ demographics and cognitive status.
Blood from younger, healthy control subjects under 55 years of age was collected for
assay optimization. Younger controls did not fast prior to blood draw.
Neuropsychological Assessment
The neuropsychological assessment consisted of the NACC UDS 2.0 (National
Alzheimer’s Coordinating Center Uniform Data Set) battery and supplemental
measures. Supplemental measures included: verbal memory, visual memory, executive
function, visuospatial ability, and global cognition. Specific supplemental tests are listed
in the following table.
176
Cognitive Domain Test Condition/Measure
Memory RAVLT Delayed Recall
WMS-4 Logical Memory Delayed Recall
WMS-4 Visual Reproduction Delayed Recall
Attention/Executive
Function
Stroop Interference
Trails A
Trails B
Language D-KEFS FAS
D-KEFS Animals
Boston Naming Test Spontaneously
Correct
Visuospatial WAIS-4 Block Design Total Correct
WMS-4 Visual Reproduction Copy
Judgment of Line Orientation Total Correct

Table 10 Supplemental Neuropsychological Measures.
In addition to the NACC-UDS 2.0, a supplemental neuropsychological battery was
administered to assess memory, attention/executive function, language, and
visuospatial ability.
RAVLT = Rey Auditory Verbal Learning Test
WMS-4 = Weschler Memory Test – 4th Edition
WAIS-4 = Weschler Adult Intelligence Scale – 4th Edition
D-KEFS = Delis-Kaplan Executive Functioning System

177
Mild Cognitive Impairment
MCI was defined as impairment on any of the following: (a) ≥ 2 tests within a
cognitive domain, (b) ≥ 3 tests across domains, or (c) an FAQ (Functional Assessment
Questionnaire) score ≥ 6. Impairment was considered as at least 1 standard deviation
below norm-referenced scores. The MCI criteria was based upon prior studies
optimizing sensitivity and specificity of diagnosis (Clark et al., 2013; Jak et al., 2009).
Participants were defined as normal cognition if they had CDR score of 0, FAQ < 6,
and do not meet the criteria for MCI.
Participant Characteristics
Participant characteristics were assessed based on demographics, medical history,
and current prescriptions. Framingham risk scores for stroke and cardiovascular
disease (office-based BMI method) were calculated accordingly (D'Agostino et al., 1994;
2008). Framingham risk scores correspond to a 10-year stroke or cardiovascular
disease event probability based on data from the Framingham heart study. The systolic
and diastolic blood pressures were calculated from the mean of arm cuff measurements
taken during neuropsychological and MRI exams.
MRI Volumetry
T1-weighted images were obtained from MRI brain scans acquired on a Siemens
Prisma 3T scanner. Volumetric measurements were calculated by segmentation of brain
regions in the FreeSurfer software package (http://freesurfer.net). Brain region volumes
were normalized to total intracranial volume (TIV).

178
Voxel-based Morphometry and Regression
Raw T1 images were processed through the voxel-based morphometry (VBM)
toolbox in SPM12 (http://www.fil.ion.ucl.ac.uk/spm), a freely available software package
for MATLAB. VBM was processed according to previously described methods
(Ashburner and Friston, 2000). Briefly, each scan was segmented into grey matter,
white matter, and CSF using SPM12’s unified segmentation procedure, warped to a
study-specific template, and spatially normalized and smoothed. Voxel-wise regression
of CD34 counts onto normalized and smoothed T1-weighted images was conducted
with and without family-wise error correction at thresholds of p < 0.05 and p < 0.001.
Model covariates included age, sex, and TIV.
Peripheral Blood Assays
Whole blood from participants (30–50 mL) was collected into K
3
EDTA tubes and
stored on ice for 3–6 hours before processing. Tubes were inverted occasionally every
30 minutes to ensure that blood constituents did not settle. Blood was processed for
flow cytometry (15 mL) and the remainder for CFU-Hill cultures. Total white blood cells
were counted on a hemacytometer using 3% acetic acid with methylene blue.
Blood for flow cytometry and culture assays was diluted 1:1 with Dulbecco's
Phosphate Buffered Saline (DPBS) + 2% fetal bovine serum (FBS). Peripheral blood
mononuclear cells (PBMCs) were isolated by density gradient centrifugation with
Histopaque 1077 (Sigma-Aldrich) or Lymphoprep (Stem Cell Technologies) in SepMate
tubes (Stem Cell Technologies). PBMCs were washed twice with DPBS + 2% FBS at
300 × g for 8 min and 120 × g for 10 min (platelet removal) at room temperature (RT).
179
Plasma Biomarkers
Plasma samples from participants were prepared according to manufacturer
instructions for Human Biomarker 40-Plex Kit (Meso Scale Discovery, MD). Peripheral
whole blood was centrifuged at 1000 × g for 10 min at 4 °C, and plasma and pellet
fractions were stored in separate polypropylene tubes at -80 °C. Prior to assay, plasma
was thawed on wet ice and centrifuged at 2000 × g for 3 min to remove particulates.
The Human Biomarker 40-Plex Kit consists of 5 electrochemiluminscence
immunoassay plates, outlined in the following table. Stem Cell Factor (SCF) was also
assayed separately by Human SCF Quantikine ELISA Kit (R&D Systems Cat #DCK00).
Plasma biomarker concentrations were reported in pg/mL.

Panel Markers
Chemokine Panel 1 Eotaxin, Eotaxin-3, IL-8, IP-10, MCP-1, MCP-4, MDC,
MIP-1α, MIP-1β, TARC
Proinflammatory Panel 1 IFN-γ, IL-10, IL-12p70, IL-13, IL-1β, IL-2, IL-4, IL-6, IL-8,
TNF-α
Cytokine Panel 1 GM-CSF, IL-12/IL-23p40, IL-15, IL-16, IL-17A, IL-1α, IL-5,
IL-7, TNF-β, VEGF-A
Angiogenesis Panel 1 FGF (basic)*, Flt-1/VEGFR-1, PlGF*, Tie-2, VEGF-A,
VEGF-C, VEGF-D
Vascular Injury Panel 2 CRP, ICAM-1, SAA, VCAM-1

Table 11 Plasma Biomarkers Measured by Immunoassay
* bFGF and PIGF were excluded due to manufacturer QA/QC issues.

180
APOE Genotyping
APOE genotyping was conducted on the blood cell pellet fraction obtained from
plasma separation. DNA was isolated from the pellet fraction using the PureLink
Genomic DNA Mini Kit (Thermo Fisher). Genotyping was conducted on isolated DNA
using the TaqMan SNP Genotyping Assay (Thermo Fisher) on an Applied Biosystems
7300 Real Time PCR System. APOE gene SNPs were assessed for dbSNP IDs
rs429358 and rs7412. Allelic discrimination was conducted using the included qPCR
software. The ApoE4 allele was designated as rs429358-C + rs7412-C.
Flow Cytometry
PBMCs (1.0 million cells) were transferred to an unstained tube, and the remainder
transferred to a stained tube. The stained tube was incubated with 5 µL of Human BD
Fc Block (BD Biosciences) for 10–15 min at RT, and 1 µL of each of the following
antibodies were added: (1) CD34-PE-Vio770 (clone: AC136, Miltenyi Biotec), (2)
CD133-VioBright FITC (clone: AC133, Miltenyi Biotec), (3) CD309-PerCP/Cy5.5 (clone:
7D4-6, BioLegend), (4) angiotensin II type 1a receptor-PE (rabbit polyclonal, Bioss), (5)
AGTR-2-APC (clone: 364805, R&D Systems), and (6) Mas1-Alexa Fluor 350 (rabbit
polyclonal, Bioss). Tubes were incubated in the dark for 30 min at 4 °C. Samples were
washed twice with 3 mL PBS + 2% FBS, centrifuged at 300 × g for 8 min, and fixed with
2% formaldehyde in PBS until analysis.
Samples were acquired on a BD LSR II flow cytometer and analyzed on FlowJo
software. Within the lymphocyte gate, 100,000 events were recorded for the blank tube
and the entire sample was recorded for the stained tube. Fluorescence compensation
181
was automatically calculated using AbC Total Antibody Compensation Bead Kit (Thermo
Fisher) that captures both mouse and rabbit antibodies. Fluorescence-minus-one (FMO)
controls were used to set positive/negative gates for each fluorochrome in the
lymphocyte gate.
Gating Strategy
Cells were gated to exclude platelets and debris (Region 1 [R1]). A pulse geometry
gate (R2) on FSC-Area vs. FSC-Height was applied to isolate single cells. Within the
lymphocyte gate (R3), single gates were drawn for CD34+ (R4), CD133+ (R5), CD309+
(R6), AT1+ (R7), AT2 (R8), and Mas (R9) above their respective FMO bounds. These
gates were sequentially applied into the lymphocyte gate to assess different cell
subsets, including: CD34+, CD34+CD133+, CD34+CD309+, CD34+CD133+CD309+,
as well as their RAS receptor expression.
CFU-Hill Assay
PBMCs were processed for the CFU-Hill assay using a commercially available kit
(Stem Cell Technologies). Briefly, PBMCs were seeded on 12-well fibronectin-coated
plates (Corning) at 2.5 × 10
6
cells/well in CFU-Hill Liquid Medium (formerly known as
EndoCult) and incubated for 2 days at 37 °C. Non-adherent cells were collected and
replated onto 24-well fibronectin-coated plates at 1.0 × 10
6
cells/well and incubated for 3
days at 37 °C. At day 5, CFU-Hill colonies in each well were scored as a central core of
round cells surrounded by radiating spindle-shaped cells (Hill et al., 2003). Functional
assays were conducted on these wells at day 5 including: senescence-associated β-
galactosidase staining and AcLDL uptake.
182
Functional Assays
Additional ex vivo functional assays were proposed, which included characterizing
these angiogenic cells on chemotaxis and Matrigel tube formation assays. However,
initial experiments demonstrated these functional assays to be either irreproducible or
infeasible under the given experimental conditions.
Senescence-associated β-galactosidase staining
Senescence-associated β-galactosidase (SA-β-gal) staining in CFU-Hill cells were
assessed using a Cellular Senescence Assay Kit (Marker Gene Technologies). At day 5
of the CFU-Hill assay, one well was fixed, washed, and incubated with 1 mg/mL X-Gal
for 16–20 hours. Cells were examined and photographed using an inverted microscope.
Only isolated spindle-shaped cells distant from central colonies were analyzed, and
cells with a distinctly blue cytoplasm (indicating β-gal activity) were counted. The % β-
gal+ cells was determined by counting at least 3 random fields (Assmus et al., 2003; Hill
et al., 2003; Lee et al., 2010).
AcLDL Uptake
One of the hallmarks of endothelial cells is the ability to take up acetylated LDL
(AcLDL). At day 5 of the CFU-Hill assay, the media from one well was replaced with
Medium 199 containing 10 µg/mL Alexa Fluor 488-conjugated AcLDL and incubated for
3–4 hours at 37 °C (Shaffer et al., 2006). Cells were washed, trypsinized, and fixed in
2% formaldehyde. Samples were acquired on a BD LSR II flow cytometer and analyzed
for % AcLDL+ cells on FlowJo software.

183
Statistical Analyses and Outcomes
Statistical analyses were performed using GraphPad Prism 6 and SPSS 24.0. Tests
were considered significant if they met a significance threshold (p value) below 0.05.
Participant characteristics were analyzed for significant differences between control
and MCI groups. Student’s t-test was conducted for continuous variables or Fisher’s
exact test for categorical variables. Outcomes were reported as mean ± standard
deviation (SD) or number of samples n (%). Continuous variables were assessed for
normality using skewness, kurtosis, and the Shapiro-Wilk test. Parameters were log10-
transformed as appropriate to satisfy the assumption of normality for parametric tests.
General linear models were conducted to adjust for covariates. Physiological
parameters such as MRI volumes were adjusted for age and sex. Neuropsychological
parameters, i.e. cognitive status, were adjusted for age, sex, and education.
Primary Outcomes: Blood-based Assays in Cognition and Brain Atrophy
The primary outcomes of this study were to evaluate: (1) whether pro-angiogenic
progenitor cell subsets & plasma biomarkers differed between control and MCI groups,
and (2) whether the levels of EPC subsets & plasma biomarkers correlated to MRI
hippocampal volumes.
Definition of Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers
Pro-angiogenic progenitor cell subsets included: CD34+, CD133+, CD309+,
CD34+CD133+, CD34+CD133+CD309+, CD34+CD309+, and CFU-Hill colony number.
Negatively selected cell subsets were also assessed, including CD34+CD133+CD309-
and CD34+CD133-CD309+. All cell subsets were normalized to 1 million cells.
184
CFU-Hill colonies were also normalized to start blood volume, calculated as: CFU-
Hill colony number / 1 million cells × (total cells after PBMC isolation / start blood
volume).
Plasma biomarkers included SCF and all measurable analytes from the MSD
Human Biomarker 40-Plex Kit.
Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers in Normal
Cognition and MCI
Initial sensitivity analyses evaluated whether there were any statistically significant
differences in pro-angiogenic progenitor cell subsets or plasma biomarkers between
control and MCI groups, without controlling for any covariates.
For sensitivity analyses, independent sample t-tests were conducted for parametric
variables and Mann-Whitney tests for non-parametric variables. Variables were
considered parametric if they passed the Shapiro-Wilk test for either the entire dataset
or within both control & MCI groups.
A general linear model was conducted to assess whether there was any difference in
pro-angiogenic progenitor cell subsets between groups after adjusting for age, sex, and
education.
Correlation of Pro-angiogenic Progenitor Cell Subsets and Plasma Biomarkers to
Brain Volumes
Multiple linear regression was conducted in SPSS to assess whether there was an
association between pro-angiogenic progenitor cell subsets or plasma biomarkers with
185
MRI hippocampal volumes, after adjusting for age as a covariate. Hippocampal volume
was calculated as the sum of both left and right hippocampi.
Hippocampal Volumetry in MCI
In the context of brain atrophy, hippocampal volumes between control and MCI
groups were evaluated. A multivariate general linear model was conducted to assess
whether there was any difference in MRI volumes between groups after adjusting for
age, sex, and education. MRI volume measures included both unilateral and summed
hippocampi measures, raw and normalized to TIV.
Vascular Senescence and Endothelial Dysfunction
We evaluated markers of vascular senescence, specifically whether senescence-
associated β-galactosidase (SA-β-gal) staining of CFU-Hill colonies differed between
control and MCI groups by t-test. It was previously reported that CACs in AD patients
exhibited higher SA-β-gal staining than their risk-factor controls.
Further, we investigated whether numbers of senescent cells were associated with
endothelial dysfunction. Specifically, we assessed whether % β-gal+ cells correlated to
plasma ICAM-1 levels, a marker of endothelial dysfunction and extravasation, by
Pearson correlation.
186
Results
Two participants were excluded due to HIV and Huntington’s disease. After
exclusion, 28 normally cognitive controls and 10 MCI participants were factored into the
study.
Participant Characteristics
Participants were grouped by cognitive status and evaluated for their demographics,
medical history, and current prescriptions. The MCI group was statistically significantly
older than the control group (75.8 ± 8.1 vs. 69.5 ± 6.4 years of age, p < 0.05).
No other factors differed significantly between the groups, including: sex, education
level, CDR, CVD, diabetes, dyslipidemia, smoking status, body mass index, blood
pressure, Framingham CVD & stroke risk scores, antihypertensive medication classes
(ACE inhibitors, ARBs, diuretics, calcium channel blockers, β-blockers), or statin use.
Increase in smoking among patients with MCI approached statistical significance (p =
0.09).

187
Control (n = 28) MCI (n = 10) p value
Age (years) 69.5 ± 6.4 75.8 ± 8.1 0.017
Male [n (%)] 11 (39) 6 (60) 0.293
Education (years) 16.1 ± 2.1 16.2 ± 4.0 0.510
CDR 0.02 ± 0.10 0.25 ± 0.27 0.096
Medical History
Hypertension 15 (54) 2 (20) 0.136
Cardiovascular Disease (CVD) 1 (4) 0 (0) 1.000
Diabetes 3 (11) 1 (10) 1.000
Dyslipidemia 16 (57) 6 (60) 1.000
Smoking 2 (7) 3 (30) 0.090
Body Mass Index 27.1 ± 5.9 26.4 ± 5.6 0.728
Systolic BP (mmHg) 131 ± 13 133 ± 14 0.652
Diastolic BP (mmHg) 79 ± 10 77 ± 9 0.541
Framingham Stroke Risk Score 9.6 ± 2.9 11.1 ± 3.4 0.193
Framingham CVD Risk Score 15.8 ± 3.8 17.4 ± 2.7 0.249
Current Medications [n (%)]
ACE Inhibitors 7 (25) 2 (20) 1.000
ARBs 4 (14) 0 (0) 0.556
Diuretics 3 (11) 0 (0) 0.552
Calcium Channel Blockers 3 (11) 0 (0) 0.552
β-Blockers 2 (7) 0 (0) 1.000
Statins 11 (39) 3 (30) 0.715

Table 12 Participant Characteristics.
Values are mean ± SD or n (%). The statistical significance between groups (p value)
was assessed by Student’s t-test for continuous variables or Fisher’s exact test for
categorical variables. CDR = Clinical Dementia Rating, BP = blood pressure, ACE =
angiotensin-converting enzyme, ARBs = angiotensin receptor blockers.

188
Detection of pro-angiogenic progenitor cell subsets and plasma biomarkers
All pro-angiogenic progenitor cell subsets were detected in all 31 blood samples
successfully processed for flow cytometry. CFU-Hill colonies were successfully obtained
for 27 samples. Neither angiogenic cell subsets processed for flow cytometry or colony-
forming assays satisfied the assumption of normality.
Plasma biomarkers were reliably detected in the vast majority of samples, except for
IL-1α and IL-1β. Of 28 processed plasma samples, only 18 and 3 samples were
detected for IL-1α and IL-1β, respectively. As a result, IL-1α and IL-1β were omitted
from further statistical analyses.
Normally distributed plasma biomarkers included: GM-CSF, ICAM-1, IL-12p70, IL-
15, IL-5, MIP-1β, TNF-α, TNF-β, VCAM-1, Flt-1, SCF, Tie-2, and VEGF-C. The rest of
the plasma biomarkers were considered non-parametric.
Primary Outcomes
Initial sensitivity analyses were conducted to evaluate whether there were any
statistically significant differences in cell subsets and plasma biomarkers between MCI
and normal cognition. However, as the MCI was significantly older than the control
group, a general linear model was conducted to assess whether there were any
differences in angiogenic cell subsets between groups after adjusting for age, sex, and
education.
Initial Sensitivity Analyses
Without adjusting for covariates, initial sensitivity analyses using the non-parametric
Mann-Whitney test revealed statistically significantly decreased progenitor cell subsets
189
CD133+ (p < 0.05), CD34+CD133+ (p < 0.05), and CD34+CD133+CD309- (p < 0.05),
and increased plasma biomarkers ICAM-1 (p < 0.05) and MIP-1α (p < 0.05) in MCI
compared to controls.

Figure 37 Decreased Angiogenic Cell Subsets in MCI.
CD133+, CD34+CD133+, and CD34+CD133+CD309- cells were found to be
significantly decreased in MCI (n = 31, * p < 0.05). Data shown are medians and
interquartile ranges. Cell subsets were normalized to 1 million cells. NC = normal
cognition, MCI = mild cognitive impairment.
NC
MCI
10
100
1000
10000
CD133+ cells
CD133+
*
NC
MCI
1
10
100
1000
CD34+CD133+ cells
CD34+CD133+
*
NC
MCI
0.1
1
10
100
1000
CD34+CD133+CD309- cells
CD34+CD133+CD309-
*
190

Figure 38 Increased Plasma ICAM-1 in MCI.
Plasma ICAM-1 was found to be significantly increased in MCI (n = 28, * p < 0.05). Data
shown are medians and interquartile ranges. NC = normal cognition, MCI = mild
cognitive impairment.

Figure 39 Increased Plasma MIP-1α in MCI.
Plasma MIP-1α was found to be significantly increased in MCI (n = 28, * p < 0.05). Data
shown are medians and interquartile ranges. NC = normal cognition, MCI = mild
cognitive impairment.
NC
MCI
0
500000
1×10
6
2×10
6
ICAM-1
Plasma ICAM-1 (pg/mL)
*
NC
MCI
0
20
40
60
80
MIP-1α
Plasma MIP-1α (pg/mL)
*
191
Reduction in CD34 and CD133 Subsets in MCI After Adjusting for Age, Sex, and
Education
All pro-angiogenic progenitor cell subsets were log10-transformed to satisfy the
assumption of normality. A univariate general linear model was conducted with cell
subsets as the predictor, cognitive status as outcome, and adjusted for age, sex, and
education as covariates.
CD34+, CD133+, and CD34+CD133+CD309- progenitor cells were statistically
significantly reduced in the MCI group, after adjusting for age, sex, and education (p <
0.05).
Other subsets were also reduced in the MCI group, but these differences did not
achieve statistical significance, including: CD309+ (p = 0.06), CD34+CD133+ (p = 0.06),
and CD34+CD133+CD309+ (p = 0.05). There was no difference in CFU-Hill colonies
between control and MCI groups.
192

Figure 40 Flow Cytometric Subsets in MCI.
CD34+ and CD133+ cells were reduced in MCI after adjusting for age, sex, and
education, n = 31, * p < 0.05. Data shown are estimated marginal means ± standard
error. Prior to adjustment, cell subsets were normalized to 1 million cells and log10-
transformed to satisfy normality assumption. NC = normal cognition, MCI = mild
cognitive impairment.

CD34+
CD133+
CD34+CD133+
CD34+CD133+CD309+
CD34+CD309+
-1
0
1
2
3
4
Adjusted Cells
NC
MCI
* * — —
193
Control (n = 22) MCI (n = 9)
Sig.
Cell Subsets Mean Std Error Mean Std Error
CD34 2.92 0.07 2.57 0.11 0.02
CD133 2.49 0.12 1.89 0.19 0.02
CD309 2.54 0.14 2.01 0.21 0.06
CD34+CD133+ 1.30 0.16 0.66 0.26 0.06
CD34+CD133+CD309+ 0.45 0.13 -0.08 0.21 0.05
CD34+CD309+ 1.54 0.13 1.10 0.20 0.09
CD34+CD133+CD309- 1.19 0.19 0.38 0.31 0.04
CD34+CD133-CD309+ 1.47 0.13 1.04 0.21 0.11
% EPCs in CD34+ -0.47 0.11 -0.71 0.18 0.29

Table 13 Flow Cytometry Subsets in Cognition.
Estimated marginal means of flow cytometry subsets after adjusting for age, sex, and
education. Prior to adjustment, cell subsets were normalized to 1 million cells and
log10-transformed to satisfy normality assumption.
Covariates appearing in the model were evaluated at the following values: Age =
71.16, Sex = .41, Education Level = 15.75.

194

Figure 41 CFU-Hill Colonies in MCI.
No significant difference was observed in CFU-Hill colonies between NC and MCI after
adjusting for age, sex, and education (n = 27). Data shown are means ± standard error.
NC = normal cognition, MCI = mild cognitive impairment.

Control (n = 19) MCI (n = 8)
Sig.
Mean Std Error Mean Std Error
CFU-Hill colonies/
1 million cells
0.94 0.12 0.74 0.21 0.45
CFU-Hill colonies/
mL start blood
0.80 0.15 0.86 0.25 0.85

Table 14 CFU-Hill Colonies in Cognition.
Estimated marginal means of CFU-Hill colonies were calculated after log10-
transformation, and adjustment for age, sex, and education.
Covariates appearing in the model are evaluated at the following values: Age =
72.22, Sex = .37, Education Level = 15.63.
per 1 million cells
per mL start blood
0.0
0.5
1.0
1.5
Adjusted CFU-Hill colonies
NC
MCI
195
Patients with higher CD34+ cells performed better on logical memory and visual
reproduction tests
Participants with higher circulating levels of CD34+, CD34+CD133+, and
CD34+CD133+CD309+ cells showed better age-adjusted memory ability. Post-hoc
multiple linear regression analyses confirmed that participants with higher CD34+ levels
exhibited better verbal memory when recall was assessed immediately, ΔR
2
= 0.10, β =
0.35, p = 0.04, and after a delay, ΔR
2
= 0.12, β = 0.38, p = 0.02, as well as when visual
memory was assessed immediately, ΔR
2
= 0.27, β = 0.58, p < 0.001. Participants with
greater circulating levels of CD34+CD133+ cells also exhibited better immediate recall
of visual information, ΔR
2
= 0.18, β = 0.43, p = 0.001. Participants with higher levels of
CD34+CD133+CD309+ cells also showed better verbal memory in the delayed recall
condition, ΔR
2
= 0.14, β = 0.39, p = 0.009, and retention of verbal information over the
delay (percent savings score), ΔR
2
= 0.18, β = 0.44, p = 0.01.
Bivariate correlations between age-adjusted scaled scores (SS) and angiogenic cell
subsets confirmed this finding in the following table.

196
Test
CD34+
(n = 33)
CD34+
CD133+
(n = 33)
CD34+
CD133+
CD309+
(n = 33)
CFU-Hill
colonies
(n = 28)
Logical Memory
Immediate
(SS)
0.46** 0.28 0.28 0.56**
Delayed
(SS)
0.40* 0.27 0.37* 0.31
Visual
Reproduction
Immediate
(SS)
0.49** 0.45** 0.38
Delayed
(SS)
0.30 0.18 0.22

Table 15 Angiogenic cell subsets positively correlated to tests of verbal and
visual memory.
Bivariate correlation coefficients (r) were obtained between age-adjusted scaled scores
(SS) and angiogenic cell subsets, * p < 0.05, ** p < 0.01. Cell subsets were normalized
to 1 million cells. Verbal and visual memory were assessed using Logical Memory and
Visual Reproduction tests from the Weschler Memory Test – 4th Edition.

CD34 cells correlated to voxel-based volumes in brain regions responsible for
visual and verbal memory
Voxel-wise regression of CD34+ cell counts onto cortical volumes was not significant
at p < 0.05 after family-wise error correction, but a significant cluster emerged at the
uncorrected p < 0.001 threshold. The significant cluster involved multiple posterior
regions implicated in visual and verbal memory deficits found in patients with AD,
197
including the lingual gyrus, posterior cingulate gyrus, and precuneus. Greater numbers
of CD34+ cells correlated to higher volumes in these regions.
Hippocampal volumes did not correlate to angiogenic cell subsets or plasma
biomarkers
Neither log10-transformed flow cytometry nor colony-forming angiogenic cell subsets
correlated to hippocampal volume, after adjusting for age. Plasma biomarkers also did
not show any significant correlation to hippocampal volume, after adjusting for age. A
bivariate correlation revealed that age was significantly correlated to hippocampal
volume (n = 35, Pearson correlation coefficient r = -0.474, p < 0.01).
Hippocampal volume was not significantly reduced in MCI group
No significant reductions in hippocampal volumes (unilateral, summed, raw values
and normalized to TIV) were observed between control and MCI groups, after adjusting
for age, sex, and education. However, there was a significant reduction in right and
summed hippocampal volumes in MCI (raw and normalized to TIV, both p < 0.05), when
evaluated on a T-test without adjusting for covariates. No significant reduction was
observed in left hippocampal volumes.
Vascular Senescence
SA-β-Gal in MCI
SA-β-gal staining was conducted on a representative well from the CFU-Hill assay,
and reliably detected in 20 of 27 CFU-Hill assays. The SA-β-gal assay was omitted if
there was no colony formation or insignificant numbers of spindle-shaped cells to score.
% β-gal+ cells were normally distributed according to the Shapiro-Wilk test.
198
Spindle-shaped cells exhibited no statistically significant difference in % β-gal+
reactivity between control and MCI groups (NC = 36.7 ± 13.9%, MCI = 47.0 ± 26.2%, n
= 20, p = 0.36).

Figure 42 SA-β-Gal Activity in MCI.
SA-β-gal activity was measured in spindle-shaped cells from the CFU-Hill assay, and
reported as % β-gal+ cells (n = 20, p = 0.36). NC = normal cognition, MCI = mild
cognitive impairment.

Correlation between SA-β-gal and ICAM-1
A bivariate correlation analysis showed that % β-Gal+ cells did not correlate with
plasma ICAM-1 levels (n = 18, p = 0.165). A multiple linear regression analysis with age
as a covariate also showed no significant correlation between % β-Gal cells as the
outcome and plasma ICAM-1 as the predictor variable (n = 18, p = 0.107).
NC
MCI
0
20
40
60
80
% SA-β-Gal+ cells
SA-β-Gal
p = 0.36
199
Discussion
CD34 deficits in MCI correlated with deficits in verbal and visual memory tests
and voxel-wise regression of MRI brain volumes involved with verbal and visual
memory
This study demonstrated that pro-angiogenic progenitor cell subsets (designated by
CD34, CD133, and CD309/VEGFR-2 expression) in circulation were reduced in
participants with MCI, and positively correlated to posterior cortical thickness. Thinning
of the posterior cingulate gyrus has been associated with the diagnosis of MCI, the
earliest clinical phase of AD (Bondi et al., 2008; Roberts et al., 2014; Schroeter et al.,
2009; Spulber et al., 2012). The reduction in these cell subsets is likely driven by a
deficit in progenitor cells, evident by the robust reductions in the parent progenitor
populations CD34+ and CD133+.
Participants with higher levels of CD34+ cells also exhibited better performance on
logical memory and visual reproduction tests, especially after immediate recall. These
findings suggest that circulating CD34+ cell levels correlate positively with verbal and
visual memory.
VBM analysis demonstrated that higher CD34+ counts significantly correlated with a
cluster of volumes in brain regions implicated in verbal and visual memory deficits,
including the precuneus, posterior cingulate gyrus, and lingual gyrus. These positive
associations between CD34+ cells and brain regions involved in verbal and visual
memory represents a significant association between a measurable blood-based
200
biomarker and some of the earliest detectable cognitive and structural deficits observed
in process of cognitive decline.
These findings reinforce other studies that have found circulating progenitor deficits
in cognitive decline, particularly those at the dementia stage. Importantly, this study
extends the findings of progenitor deficit to an earlier stage of cognitive decline, MCI.
What is truly unique about this study is that these progenitor cell deficits also correlated
to neuropsychological and structural-based MRI markers involved in verbal and visual
memory.
Identity, significance, and association of progenitor cells in AD
Prior studies have found CD34+CD133+ cells and CFU-EPCs to be reduced and
dysfunctional in AD (Kong et al., 2011; Lee et al., 2009; 2010). These cells (i.e. CFU-
EPC, CFU-Hill, CD34+CD133+/-VEGFR-2+/- cells) have been more recently defined as
pro-angiogenic progenitor cells based on their progenitor markers (CD34 & CD133),
proliferative ability, and angiogenic potential (Basile and Yoder, 2014). These progenitor
cells have been shown to exhibit pro-angiogenic potential by forming colonies in
endothelial growth medium or by co-expressing endothelial markers (e.g.
CD309/VEGFR-2). Therefore, it is likely that these prior observations are likely related
to the deficits in pro-angiogenic progenitor cells.
Perhaps one of the most difficult aspects in determining the role of progenitor cells in
AD is ascertaining whether these associations are truly causative. It is complicated by
the fact that a lot of these associations may be temporally related or distinct, as we see
a decrease in CD34+ cells early on in AD (Kong et al., 2011; Maler et al., 2006), while
201
another study found an increase in CD34+ cells in mid-to-severe AD (Stellos et al.,
2010). This later increase in mobilized CD34+ cells may be a response to injury in the
brain.
Role of these pro-angiogenic progenitor cells in cognitive decline
Despite the significance of these pro-angiogenic progenitor cells in cognitive decline
in respect to MCI and AD, the exact role of these cells is still undetermined. It is not
certain whether these cells are a cause or consequence of cognitive decline, and
whether they specifically relate to AD pathology or another age-dependent mechanism.
Further studies using longitudinal design and markers of AD pathophysiology (e.g. CSF
Aβ or tau) may be needed to elucidate their role. It will be of great interest whether
longitudinal studies and greater participant recruitment can show whether these cells
can predict conversion to MCI or AD.
Within the context of vascular regeneration, pro-angiogenic progenitor cell subsets
represented by CD34, CD133, and CD309 markers and CFU-Hill colonies do not fit the
criteria for a true EPC. An EPC is defined as a single cell with high proliferative capacity
that gives rise to endothelial cells. Instead, these pro-angiogenic progenitor cells likely
exert a paracrine angiogenic effect through trophic factors.
There are limitations as to what we can observe clinically in regards to progenitor
cell proliferation, mobilization, homing, and differentiation. Mobilization of CD34+ cells
into the peripheral bloodstream does not indicate where they will end up ultimately, and
whether they differentiate into functional parenchymal cells. These progenitor cells may
play a role through transdifferentiation into neuroglia. A recent study has demonstrated
202
that transplantation of ex vivo culture of BM-EPCs showed successful incorporation into
murine brain parenchyma and expression of neuronal markers (Safar et al., 2014).
Vascular risk and participant population
It should be noted that these findings do not appear to be related to vascular risk or
disease. This is of great consideration, since prior studies have found that CFU-Hill
colonies negatively correlated to cardiovascular risk and vascular diseases (Hill et al.,
2003). Our dataset represents a healthy patient population; participants were living
independently, ambulatory, and community dwelling. Participants were free of stroke
and cardiovascular disease at study admission. Thus, the study outcomes represent
findings independent of confounding diseases associated with pro-angiogenic
progenitor cells.
Despite the small sample size, this significant finding correlated with both
neuropsychological & MRI measures
While there have been similar studies investigating pro-angiogenic progenitor cell
deficits in cognitive decline, this is the first study investigating these cell subsets in
conjunction with an extensive neuropsychological battery and structural MRI findings.
Matched longitudinal studies will be increasingly important for validating these findings,
as they can show cognitive performance over time along with matched MRI and
progenitor findings.
The most significant confounder in this clinical study was that the MCI group was
significantly older than the NC group. Since age is the most significant risk factor in
cognitive decline and dementia, it was crucial to control for age to ascertain clinical
203
findings independent of age risk. Thus, the use of statistical analyses to control for
covariates, e.g. ANCOVA, was necessary to assess the effects of pro-angiogenic
progenitor cells in MCI after adjusting for age, sex, and education. The ongoing,
longitudinal nature of this study will continue to recruit participants to increase statistical
power and permit other robust statistical analyses such as risk-factor matching and
within-subject matching.
Vascular senescence
Despite the evidence by Lee at al. that the progeny of CFU-EPCs exhibit increased
cellular senescence (SA-β-Gal staining) in AD patients, we did not find evidence of this
in MCI even without controlling for age. Both NC & MCI participants in this study
exhibited similar levels of SA-β-Gal+ staining to the risk-factor controls in the Lee study,
while the AD group was significantly higher (VaSC NC = 36.7 ± 13.9%, VaSC MCI =
47.0 ± 26.2%, Lee RF controls = 33.2 ± 23.5%, Lee AD = 73.5 ± 21.2%). It is possible
that vascular senescence associated with neurodegenerative processes at the
dementia stage are not yet present at the MCI stage.
Other factors worth considering is that we used the CFU-Hill assay, instead of the
CAC assay in the Lee study. CFU-Hill colonies give rise to distinctly spindle-shaped
cells, whereas cells in the CAC assay adopt a cobblestone-like morphology. While the
SA-β-Gal method is a commonly used method to evaluate cellular senescence, its utility
in colony-forming assays has limitations. This method evaluates the senescence of
progeny from proliferating colonies. CFU-Hill colonies with high proliferative capacity
yield high numbers of spindle-shaped cells, which makes enumeration of senescent
204
cells possible. In contrast, colonies with low proliferative capacity and few cells makes
senescent characterization unreliable. Thus, the results of this senescence assay were
skewed away from participants with low colony number or low proliferative capacity.
This was evident in this dataset, as a quarter of the participant samples did not yield
reliable enumeration of SA-β-Gal cells.
Sensitivity analyses revealed that plasma ICAM-1 was significantly elevated in the
MCI group compared to the NC group, without adjusting for age. ICAM-1 is a marker of
endothelial dysfunction, and has a role in cell-cell adhesion and leukocyte
extravasation. Plasma biomarker studies have previously found that ICAM-1 to be
significantly elevated in AD, suggesting that endothelial dysfunction may be implicated
in AD pathogenesis. Our finding that ICAM-1 was also elevated in the MCI group implies
that endothelial dysfunction may be present early in cognitive decline.
However, this finding is confounded by the fact that the MCI group was significantly
older than the NC group. Plasma ICAM-1 finding was not significant after age
adjustment between the NC vs. MCI groups. It is unclear whether age adjustment is
appropriate; ICAM-1 may instead be an indicator of underlying cardiovascular risk, and
cumulative cardiovascular risk increases with age. One study suggests that plasma
ICAM-1 levels may be predictive of future coronary events, independent of age (Luc et
al., 2003). Further, plasma ICAM-1 levels did not correlate to SA-β-Gal+ cells in our
dataset, despite my initial hypothesis and expectations that they would, since vascular
senescence contributes to endothelial dysfunction.
205
Vascular senescence may be exacerbated at later stages in AD pathogenesis. At the
MCI stage, according to our data, vascular senescence appears to play a small role
since we have shown little-to-no difference in age-adjusted senescence markers
between NC and MCI. However, this study into vascular senescence is limited due to
the small sample size. Vascular senescence did not appear to play a large role at this
stage in MCI, but it may correlate with other subclinical MRI findings such as white
matter lesions. Further work and more statistical power will be necessary to elucidate
the role of vascular senescence in cognitive decline.
206
Summary
In summary, significant deficits in pro-angiogenic progenitor cell subsets (CD34,
CD133, CD309/VEGFR-2) were observed in MCI participants after adjusting for age,
sex, and education. Higher numbers of CD34+ cells correlated with better performance
on verbal and visual memory tests. Structural MRI studies further corroborated these
findings, in which higher CD34+ cells also strongly correlated to volumes in brain
regions implicated in verbal and visual memory deficits, including the precuneus,
posterior cingulate gyrus, and lingual gyrus.
We have previously observed that CD34+CD133+ cells positively correlated to CFU-
Hill colonies, suggesting that they may derive from a similar population of progenitor
cells. Prior studies have shown that both CD34+CD133+ and CFU-EPCs were
diminished in AD (Kong et al., 2011; Lee et al., 2009). Interestingly, we only observed a
deficit in CD34+CD133+ cells, while CFU-Hill colony numbers were preserved in MCI.
This may reflect a diminished number of circulating progenitor cells, while the function of
these cells may still be preserved in MCI. Preservation of function may be evident by
colony formation and lack of senescence. Further along cognitive decline, both
progenitor cell number and function may be compromised at the dementia stage.
Taken altogether, circulating pro-angiogenic progenitor cells may be protective in
cognitive decline, and reduction in progenitor cell numbers occur in the earliest
detectable clinical phase of MCI.
207
Chapter 6: Conclusions and Future Directions
In conclusion, a great deal of knowledge has been gained to elucidate the protective
role of the RAS and angiogenic cells in cognitive decline. Much of this work stands upon
the collective shoulders of giants—scientists, clinicians, patients, caregivers, and loved
ones—as we all strive to better understand the diseases of old age and to improve the
human condition. Controlling modifiable vascular risk factors such as midlife
hypertension may help stave cognitive decline. Beyond blood pressure control,
modulating the central actions of the RAS using pharmacological agents such as
centrally active ARBs or Mas agonists may confer additional cognitive protection,
possibly through increasing clearance of pathological Aβ. As further research comes to
light, neuroprotective RAS agents may emerge as safe and effective therapeutics for
AD. Further, circulating progenitor cells may be protective in the early stages of
cognitive decline, and may serve as a useful diagnostic, biomarker, or point of
intervention. Pivotal future directions may aid in the translation of these findings to
clinical use.
As a small peptide drug with a short half-life, Ang-(1–7) must be injected chronically
and subcutaneously. Systemic administration likely only allows a modicum of intact
peptide to reach the brain. Bypassing the BBB represents a major hurdle that plagues
the development of therapies for neurodegenerative diseases such as AD. Synthetic
Mas agonists that penetrate the BBB may be a viable path towards activating the
neuroprotective RAS pathways.
208
Alternatively, intranasal administration is an intriguing route that may aid in delivering
neuroprotective RAS agents such as Ang-(1–7) to the brain. Proper intranasal
formulations and dosage studies would be required, although this represents a new and
exciting future direction of research. Intranasal delivery systems have already been
shown to be feasible and translatable from rodents to humans, as much of this work has
been pioneered by the development of intranasal insulin. In addition, intranasal delivery
may avoid undesired peripheral hypotensive effects. As proof of concept, sub-
antihypertensive doses of intranasal losartan have already been shown to significantly
reduce amyloid plaques. Whether this mechanism involves activation of ACE2/Ang-(1–
7)/Mas remains to be seen.
While candesartan and Ang-(1–7) significantly reduced AD pathology in 3xTg-AD
mice, effects on cognitive outcomes were not established. Future studies to replicate
these results should consider phenotyping spatial memory using either the Morris Water
Maze or Barnes maze. In addition, while protective RAS agents may prevent amyloid
deposition when started at an early age, it is unknown whether these protective RAS
agents can clear Aβ in aged animals with significant amyloid load. Paradigms of
initiating treatment at young or old age distinguishes a preventive role versus a
therapeutic role. It may be prudent to evaluate the efficacy of protective RAS agents in
5xFAD mice. The 5xFAD model incorporates 5 FAD mutations, which results in a severe
and aggressive model of Aβ deposition that occurs much earlier than all other existing
animal models.
209
The significance and role of circulating progenitor cells in cognitive decline is only
now beginning to surface. We have shown that circulating progenitor cells were
diminished in MCI, and these deficits correlated to structures and cognitive domains
affected early in cognitive decline. This suggests circulating progenitor cells may be
protective in the early stages of cognitive decline, but their exact role and relative
contribution to cognitive decline is unknown. As this clinical study is ongoing and
longitudinal in nature, further participant recruitment will shed light on this issue.
Matched participant samples over time will allow investigations into the relative
contribution of these circulating progenitor cells to cognitive decline, specifically
conversion rates to MCI or AD. It is also unknown whether genetic risk factors such as
ApoE genotype plays a role.
The exact role of these progenitor cells, and their angiogenic potential is uncertain.
Future conditioned media studies may clarify their angiogenic potential. Conducting a
Matrigel assay with conditioned media may yield functional information about
angiogenic potential. Another potential future direction is to further investigate the role of
mobilized progenitor cell mobilization in animal models (Shin et al., 2011).
Several avenues of resources may expand the line of research of circulating
progenitor cells in cognitive decline. Collection of clinical samples represents a
significant logistical bottleneck and challenge. Biobanks represent one underutilized
facet of biomedical research. Biobanks collect and preserve tissues, including blood and
brain samples, for future medical research. It may be worth investigating whether
biobanked blood cells yield viable progenitor cells to bolster sample size, and possibly
210
expanding potential paths for further exploratory studies. Along the same line of thought,
it may be worth considering biobanking or preserving participant replicate samples to
revisit for future studies.
211
References
Ahrens, I., Domeij, H., Topcic, D., Haviv, I., Merivirta, R.-M., Agrotis, A., Leitner,
E., Jowett, J.B., Bode, C., Lappas, M., et al. (2011). Successful in vitro expansion
and differentiation of cord blood derived CD34+ cells into early endothelial
progenitor cells reveals highly differential gene expression. PLoS ONE 6,
e23210.
Albiston, A.L., Diwakarla, S., Fernando, R.N., Mountford, S.J., Yeatman, H.R.,
Morgan, B., Pham, V., Holien, J.K., Parker, M.W., Thompson, P.E., et al. (2011).
Identification and development of specific inhibitors for insulin-regulated
aminopeptidase as a new class of cognitive enhancers. British Journal of
Pharmacology 164, 37–47.
Alenina, N., Xu, P., Rentzsch, B., Patkin, E.L., and Bader, M. (2008). Genetically
altered animal models for Mas and angiotensin-(1-7). Experimental Physiology
93, 528–537.
Asahara, T., Murohara, T., Sullivan, A., Silver, M., van der Zee, R., Li, T.,
Witzenbichler, B., Schatteman, G., and Isner, J.M. (1997). Isolation of putative
progenitor endothelial cells for angiogenesis. Science 275, 964–967.
Ashburner, J., and Friston, K.J. (2000). Voxel-Based Morphometry—The
Methods. NeuroImage 11, 805–821.
Assmus, B., Urbich, C., Aicher, A., Hofmann, W.K., Haendeler, J., Rössig, L.,
Spyridopoulos, I., Zeiher, A.M., and Dimmeler, S. (2003). HMG-CoA reductase
inhibitors reduce senescence and increase proliferation of endothelial progenitor
cells via regulation of cell cycle regulatory genes. Circ. Res. 92, 1049–1055.
Association, A.D. (2008). Standards of Medical Care in Diabetes—2008.
Diabetes Care 31, S12–S54.
Attwell, D., and Iadecola, C. (2002). The neural basis of functional brain imaging
signals. Trends in Neurosciences.
Badorff, C., Brandes, R.P., Popp, R., Rupp, S., Urbich, C., Aicher, A., Fleming, I.,
Busse, R., Zeiher, A.M., and Dimmeler, S. (2003). Transdifferentiation of blood-
derived human adult endothelial progenitor cells into functionally active
cardiomyocytes. Circulation 107, 1024–1032.
Bahlmann, F.H., de Groot, K., Mueller, O., Hertel, B., Haller, H., and Fliser, D.
(2005). Stimulation of endothelial progenitor cells: a new putative therapeutic
212
effect of angiotensin II receptor antagonists. 45, 526–529.
Barrett, A.J., Rawlings, N.D., and Woessner, J.F. (2012). Handbook of
Proteolytic Enzymes (Academic Press).
Basile, D.P., and Yoder, M.C. (2014). Circulating and tissue resident endothelial
progenitor cells. J. Cell. Physiol. 229, 10–16.
Bastos-Leite, A.J., van Waesberghe, J.H., Oen, A.L., van der Flier, W.M.,
Scheltens, P., and Barkhof, F. (2006). Hippocampal sulcus width and cavities:
comparison between patients with Alzheimer disease and nondemented elderly
subjects. AJNR Am J Neuroradiol 27, 2141–2145.
Becker, L.K., Etelvino, G.M., Walther, T., Santos, R.A.S., and Campagnole-
Santos, M.J. (2007). Immunofluorescence localization of the receptor Mas in
cardiovascular-related areas of the rat brain. Am. J. Physiol. Heart Circ. Physiol.
293, H1416–H1424.
Benicky, J., Hafko, R., Sánchez-Lemus, E., Aguilera, G., and Saavedra, J.M.
(2012). Six commercially available angiotensin II AT1 receptor antibodies are
non-specific. Cell. Mol. Neurobiol. 32, 1353–1365.
Bernstein, K.E., Koronyo, Y., Salumbides, B.C., Sheyn, J., Pelissier, L., Lopes,
D.H.J., Shah, K.H., Bernstein, E.A., Fuchs, D.-T., Yu, J.J.-Y., et al. (2014).
Angiotensin-converting enzyme overexpression in myelomonocytes prevents
Alzheimer's-like cognitive decline. J. Clin. Invest. 124, 1000–1012.
Bevins, R.A., and Besheer, J. (2006). Object recognition in rats and mice: a one-
trial non-matching-to-sample learning task to study“recognition memory.” Nat
Protoc 1, 1306–1311.
Birbrair, A., and Frenette, P.S. (2016). Niche heterogeneity in the bone marrow.
Ann. N. Y. Acad. Sci. 1370, 82–96.
Blaufox, M.D., Lee, H.B., Davis, B., Oberman, A., Wassertheil-Smoller, S., and
Langford, H. (1992). Renin predicts diastolic blood pressure response to
nonpharmacologic and pharmacologic therapy. Jama 267, 1221–1225.
Bondi, M.W., Jak, A.J., Delano-Wood, L., Jacobson, M.W., Delis, D.C., and
Salmon, D.P. (2008). Neuropsychological contributions to the early identification
of Alzheimer's disease. Neuropsychol Rev 18, 73–90.
Bortolato, M., Frau, R., Bini, V., Luesu, W., Loriga, R., Collu, M., Gessa, G.L.,
Ennas, M.G., and Castelli, M.P. (2010). Methamphetamine neurotoxicity
213
increases brain expression and alters behavioral functions of CB₁ cannabinoid
receptors. J Psychiatr Res 44, 944–955.
Bouhlel, M.A., Derudas, B., Rigamonti, E., Dièvart, R., Brozek, J., Haulon, S.,
Zawadzki, C., Jude, B., Torpier, G., Marx, N., et al. (2007). PPARgamma
activation primes human monocytes into alternative M2 macrophages with anti-
inflammatory properties. Cell Metabolism 6, 137–143.
Braak, H., and Braak, E. (1991). Neuropathological stageing of Alzheimer-related
changes. Acta Neuropathol 82, 239–259.
Braak, H., Alafuzoff, I., Arzberger, T., Kretzschmar, H., and Del Tredici, K.
(2006). Staging of Alzheimer disease-associated neurofibrillary pathology using
paraffin sections and immunocytochemistry.
Braff, D.L., Geyer, M.A., and Swerdlow, N.R. (2001). Human studies of prepulse
inhibition of startle: normal subjects, patient groups, and pharmacological
studies. Psychopharmacology (Berl.) 156, 234–258.
Breunig, J.J., Guillot-Sestier, M.-V., and Town, T. (2013). Brain injury,
neuroinflammation and Alzheimer's disease. Front Aging Neurosci 5.
Britschgi, M., and Wyss-Coray, T. (2009). Blood Protein Signature for the Early
Diagnosis of Alzheimer Disease. Arch Neurol 66, 161–165.
Buccafusco, J.J., Bryan, K.J., Lee, H.-G., Perry, G., Smith, M.A., and Casadesus,
G. (2009). Transgenic Mouse Models of Alzheimer’s Disease: Behavioral Testing
and Considerations (Boca Raton (FL): CRC Press).
Callera, G.E., Antunes, T.T., Correa, J.W., Moorman, D., Gutsol, A., He, Y., Cat,
A.N.D., Briones, A.M., Montezano, A.C., Burns, K.D., et al. (2016). Differential
renal effects of candesartan at high and ultra-high doses in diabetic mice-
potential role of the ACE2/AT2R/Mas axis. Biosci. Rep. 36, e00398–e00398.
Cardoso, C.C., Alenina, N., Ferreira, A.J., Qadri, F., Lima, M.P., Gross, V.,
Todiras, M., Pesquero, J.B., Pesquero, J.L., and Bader, M. (2010). Increased
blood pressure and water intake in transgenic mice expressing rat tonin in the
brain. Biol. Chem. 391, 435–441.
Caruso, D., Barron, A.M., Brown, M.A., Abbiati, F., Carrero, P., Pike, C.J.,
Garcia-Segura, L.M., and Melcangi, R.C. (2013). Age-related changes in
neuroactive steroid levels in 3xTg-AD mice. Neurobiol. Aging 34, 1080–1089.
Chai, S.Y. (2013). Angiotensin-Converting Enzyme in the Human Brain. In
214
Receptors in the Human Nervous System, (Receptors in the Human Nervous
System).
Chai, S.Y., Bastias, M.A., Clune, E.F., Matsacos, D.J., Mustafa, T., Lee, J.H.,
McDowall, S.G., Mendelsohn, F.A.O., Albiston, A.L., and Paxinos, G. (2000).
Distribution of angiotensin IV binding sites (AT4 receptor) in the human forebrain,
midbrain and pons as visualised by in vitro receptor autoradiography. Journal of
Chemical Neuroanatomy 20, 339–348.
Chin, J. (2011). Selecting a Mouse Model of Alzheimer’s Disease. In Alzheimer's
Disease and Frontotemporal Dementia, (Totowa, NJ: Humana Press), pp. 169–
189.
Clark, L.R., Delano-Wood, L., Libon, D.J., McDonald, C.R., Nation, D.A., Bangen,
K.J., Jak, A.J., Au, R., Salmon, D.P., and Bondi, M.W. (2013). Are Empirically-
Derived Subtypes of Mild Cognitive Impairment Consistent with Conventional
Subtypes? J Int Neuropsychol Soc 19, 635–645.
Clinton, L.K., Billings, L.M., Green, K.N., Caccamo, A., Ngo, J., Oddo, S.,
McGaugh, J.L., and LaFerla, F.M. (2007). Age-dependent sexual dimorphism in
cognition and stress response in the 3xTg-AD mice. Neurobiol. Dis. 28, 76–82.
Corbett, A., Pickett, J., Burns, A., Corcoran, J., Dunnett, S.B., Edison, P., Hagan,
J.J., Holmes, C., Jones, E., Katona, C., et al. (2012). Drug repositioning for
Alzheimer's disease. Nat Rev Drug Discov 11, 833–846.
Culman, J., Blume, A., Gohlke, P., and Unger, T. (2002). The renin-angiotensin
system in the brain: possible therapeutic implications for AT1-receptor blockers. J
Hum Hypertens 16, S64–S70.
D'Agostino, R.B., Wolf, P.A., Belanger, A.J., and Kannel, W.B. (1994). Stroke risk
profile: adjustment for antihypertensive medication. The Framingham Study.
Stroke 25, 40–43.
D'Agostino, R.B., Vasan, R.S., Pencina, M.J., Wolf, P.A., Cobain, M., Massaro,
J.M., and Kannel, W.B. (2008). General cardiovascular risk profile for use in
primary care: the Framingham Heart Study. Circulation 117, 743–753.
Danielyan, L., Klein, R., Hanson, L.R., Buadze, M., Schwab, M., Gleiter, C.H.,
and Frey, W.H., II (2010). Protective effects of intranasal losartan in the APP/PS1
transgenic mouse model of Alzheimer disease. Rejuvenation Res 13, 195–201.
Davies, N.M., Kehoe, P.G., Ben-Shlomo, Y., and Martin, R.M. (2011).
215
Associations of anti-hypertensive treatments with Alzheimer's disease, vascular
dementia, and other dementias. J. Alzheimers Dis. 26, 699–708.
Davisson, R.L. (2003). Physiological genomic analysis of the brain renin-
angiotensin system. Am J Physiol Regul Integr Comp Physiol 285, R498–R511.
de Kloet, A.D., Liu, M., Rodriguez, V., Krause, E.G., and Sumners, C. (2015).
Role of neurons and glia in the CNS actions of the renin-angiotensin system in
cardiovascular control. Am J Physiol Regul Integr Comp Physiol
ajpregu.00078.2015–ajpregu.00078.2066.
De Mota, N., Iturrioz, X., Claperon, C., Bodineau, L., Fassot, C., Roques, B.P.,
Palkovits, M., and Llorens-Cortes, C. (2008). Human brain aminopeptidase A:
biochemical properties and distribution in brain nuclei. Journal of Neurochemistry
106, 416–428.
Deacon, R.M.J., and Rawlins, J.N.P. (2006). T-maze alternation in the rodent.
Nat Protoc 1, 7–12.
Denny, J.B., Polan-Curtain, J., Wayner, M.J., and Armstrong, D.L. (1991).
Angiotensin II blocks hippocampal long-term potentiation. Brain Res. 567, 321–
324.
Desai, A., Glaser, A., Liu, D., Raghavachari, N., Blum, A., Zalos, G., Lippincott,
M., McCoy, J.P., Munson, P.J., Solomon, M.A., et al. (2009). Microarray-based
characterization of a colony assay used to investigate endothelial progenitor cells
and relevance to endothelial function in humans. Arterioscler Thromb Vasc Biol
29, 121–127.
Dhaunsi, G.S., Yousif, M.H.M., Akhtar, S., Chappell, M.C., Diz, D.I., and Benter,
I.F. (2010). Angiotensin-(1–7) prevents diabetes-induced attenuation in PPAR-γ
and catalase activities. European Journal of Pharmacology 638, 108–114.
Di Pollina, L., Mulligan, R., Juillerat Van der Linden, A., Michel, J.P., and Gold, G.
(2000). Cognitive impairment in polycythemia vera: partial reversibility upon
lowering of the hematocrit. Eur. Neurol. 44, 57–59.
Doobay, M.F., Talman, L.S., Obr, T.D., Tian, X., Davisson, R.L., and Lazartigues,
E. (2007). Differential Expression of Neuronal ACE2 in Transgenic Mice with
Overexpression of the Brain Renin-angiotensin System. Am J Physiol Regul
Integr Comp Physiol 292, R373–R381.
Droste, D.W., Ritter, M.A., and Dittrich, R. (2003). Arterial hypertension and
216
ischaemic stroke. Acta Neurologica ….
Duchemin, S., Belanger, E., Wu, R., Ferland, G., and Girouard, H. (2013).
Chronic perfusion of angiotensin II causes cognitive dysfunctions and anxiety in
mice. Physiology & Behavior 109, 63–68.
Durik, M., Pessôa, B.S., and Roks, A.J.M. (2012). The renin–angiotensin system,
bone marrow and progenitor cells. Clinical Science 123, 205–223.
Ellefson, D.D., Dizerega, G.S., Espinoza, T., Roda, N., Maldonado, S., and
Rodgers, K.E. (2004). Synergistic effects of co-administration of angiotensin 1-7
and Neupogen on hematopoietic recovery in mice. Cancer Chemother.
Pharmacol. 53, 15–24.
Endtmann, C., Ebrahimian, T., Czech, T., Arfa, O., Laufs, U., Fritz, M.,
Wassmann, K., Werner, N., Petoumenos, V., Nickenig, G., et al. (2011).
Angiotensin II impairs endothelial progenitor cell number and function in vitro and
in vivo: implications for vascular regeneration. 58, 394–403.
Erbe, D.V., Gartrell, K., Zhang, Y.-L., Suri, V., Kirincich, S.J., Will, S., Perreault,
M., Wang, S., and Tobin, J.F. (2006). Molecular activation of PPARgamma by
angiotensin II type 1-receptor antagonists. Vascul. Pharmacol. 45, 154–162.
Eriksson, P.S., Perfilieva, E., Björk-Eriksson, T., Alborn, A.-M., Nordborg, C.,
Peterson, D.A., and Gage, F.H. (1998). Neurogenesis in the adult human
hippocampus. Nat. Med. 4, 1313–1317.
Estes, M.L., Mund, J.A., Mead, L.E., Prater, D.N., Cai, S., Wang, H., Pollok, K.E.,
Murphy, M.P., An, C.S.T., Srour, E.F., et al. (2010). Application of Polychromatic
Flow Cytometry to Identify Novel Subsets of Circulating Cells with Angiogenic
Potential. Cytometry A 77, 831–839.
Fadini, G.P., Losordo, D., and Dimmeler, S. (2012). Critical reevaluation of
endothelial progenitor cell phenotypes for therapeutic and diagnostic use. Circ.
Res. 110, 624–637.
Faraci, F.M., and Heistad, D.D. (1998). Regulation of the Cerebral Circulation:
Role of Endothelium and Potassium Channels. Physiol. Rev. 78, 53–97.
Ferrington, L., Palmer, L.E., and Love, S. (2012). Angiotensin II-inhibition: effect
on Alzheimer’s pathology in the aged triple transgenic mouse. American Journal
of ….
Ferrington, L., Miners, J.S., Palmer, L.E., Bond, S.M., Povey, J.E., Kelly, P.A.,
217
Love, S., Horsburgh, K.J., and Kehoe, P.G. (2011). Angiotensin II-inhibiting drugs
have no effect on intraneuronal Aβ or oligomeric Aβ levels in a triple transgenic
mouse model of Alzheimer's disease. Am J Transl Res 3, 197–208.
Fierstra, J., Sobczyk, O., Battisti Charbonney, A., Mandell, D.M., Poublanc, J.,
Crawley, A.P., Mikulis, D.J., Duffin, J., and Fisher, J.A. (2013). Measuring
cerebrovascular reactivity: what stimulus to use? J. Physiol. (Lond.) 591, 5809–
5821.
Fitzsimons, J.T. (1998). Angiotensin, thirst, and sodium appetite. Physiol. Rev.
78, 583–686.
Freund, M., Walther, T., and Bohlen und Halbach, von, O. (2012).
Immunohistochemical localization of the angiotensin-(1-7) receptor Mas in the
murine forebrain. Cell Tissue Res. 348, 29–35.
Frisoni, G.B., Fox, N.C., Jack, C.R., Jr, Scheltens, P., and Thompson, P.M.
(2010). The clinical use of structural MRI in Alzheimer disease. Nat Rev Neurol 6,
67–77.
Gao, J., Marc, Y., Iturrioz, X., Leroux, V., Balavoine, F., and Llorens-Cortes, C.
(2014). A new strategy for treating hypertension by blocking the activity of the
brain renin–angiotensin system with aminopeptidase A inhibitors. Clinical
Science 127, 135–148.
Gao, Y., O'Caoimh, R., Healy, L., Kerins, D.M., Eustace, J., Guyatt, G., Sammon,
D., and Molloy, D.W. (2013). Effects of centrally acting ACE inhibitors on the rate
of cognitive decline in dementia. BMJ Open 3, e002881.
García-Mesa, Y., López-Ramos, J.C., Giménez-Llort, L., Revilla, S., Guerra, R.,
Gruart, A., LaFerla, F.M., Cristòfol, R., Delgado-García, J.M., and Sanfeliu, C.
(2011). Physical exercise protects against Alzheimer's disease in 3xTg-AD mice.
J. Alzheimers Dis. 24, 421–454.
Geyer, M.A., and Dulawa, S.C. (2003). Assessment of murine startle reactivity,
prepulse inhibition, and habituation. Curr Protoc Neurosci Chapter 8, Unit8.17.
Geyer, M.A., and Swerdlow, N.R. (1998). Measurement of startle response,
prepulse inhibition, and habituation. Current Protocols in Neuroscience.
Girouard, H., Lessard, A., Capone, C., Milner, T.A., and Iadecola, C. (2008). The
neurovascular dysfunction induced by angiotensin II in the mouse neocortex is
sexually dimorphic. Am. J. Physiol. Heart Circ. Physiol. 294, H156–H163.
218
Girouard, H., and Iadecola, C. (2006). Neurovascular coupling in the normal
brain and in hypertension, stroke, and Alzheimer disease. J. Appl. Physiol. 100,
328–335.
Gohlke, P., Kügelgen, Von, S., Jürgensen, T., Kox, T., Rascher, W., Culman, J.,
and Unger, T. (2002). Effects of orally applied candesartan cilexetil on central
responses to angiotensin II in conscious rats. J. Hypertens. 20, 909–918.
Goon, P.K.Y., Lip, G.Y.H., Boos, C.J., Stonelake, P.S., and Blann, A.D. (2006).
Circulating Endothelial Cells, Endothelial Progenitor Cells, and Endothelial
Microparticles in Cancer. Neoplasia (New York, N.Y.) 8, 79–88.
Gorelick, P.B., Scuteri, A., Black, S.E., DeCarli, C., Greenberg, S.M., Iadecola,
C., Launer, L.J., Laurent, S., Lopez, O.L., Nyenhuis, D., et al. (2011). Vascular
contributions to cognitive impairment and dementia: a statement for healthcare
professionals from the american heart association/american stroke association.
Stroke 42, 2672–2713.
Gottesman, R.F., Schneider, A.L.C., Albert, M., Alonso, A., Bandeen-Roche, K.,
Coker, L., Coresh, J., Knopman, D., Power, M.C., Rawlings, A., et al. (2014).
Midlife hypertension and 20-year cognitive change: the atherosclerosis risk in
communities neurocognitive study. - PubMed - NCBI. JAMA Neurol 71, 1218–
1227.
Goussetis, E., Theodosaki, M., Paterakis, G., Peristeri, J., Petropoulos, D., Kitra,
V., Papassarandis, C., and Graphakos, S. (2000). A functional hierarchy among
the CD34+ hematopoietic cells based on in vitro proliferative and differentiative
potential of AC133+CD34(bright) and ... - PubMed - NCBI. Journal of
Hematotherapy & Stem Cell Research 9, 827–840.
Grigsby, J., Kaye, K., Shetterly, S.M., Baxter, J., Morgenstern, N.E., and
Hamman, R.F. (2002). Prevalence of Disorders of Executive Cognitive
Functioning among the Elderly: Findings from the San Luis Valley Health and
Aging Study. Neuroepidemiology 21, 213–220.
Grisar, J., Aletaha, D., Steiner, C.W., Kapral, T., Steiner, S., Seidinger, D.,
Weigel, G., Schwarzinger, I., Wolozcszuk, W., Steiner, G., et al. (2005).
Depletion of endothelial progenitor cells in the peripheral blood of patients with
rheumatoid arthritis. Circulation 111, 204–211.
Hafko, R., Villapol, S., Nostramo, R., Symes, A., Sabban, E.L., Inagami, T., and
Saavedra, J.M. (2013). Commercially available angiotensin II At₂ receptor
antibodies are nonspecific. PLoS ONE 8, e69234.
219
Hajjar, I. (2012). Impact of Angiotensin Receptor Blockers on Alzheimer Disease
Neuropathology in a Large Brain Autopsy Series. Arch Neurol 69, 1632.
Harwood, D.G., and Sultzer, D.L. (2000). Impaired insight in Alzheimer disease:
association with cognitive deficits, psychiatric symptoms, and behavioral
disturbances. Cognitive and Behavioral ….
Hay, M., Vanderah, T.W., Samareh-Jahani, F., Constantopoulos, E., Uprety,
A.R., Barnes, C.A., and Konhilas, J. (2017). Cognitive Impairment in Heart
Failure: A Protective Role for Angiotensin-(1-7). Behav. Neurosci. 131, 99–114.
Healy, D.P., and Wilk, S. (1993). Localization of immunoreactive glutamyl
aminopeptidase in rat brain. II. Distribution and correlation with angiotensin II.
Brain Res. 606, 295–303.
Hebert, L.E., Weuve, J., Scherr, P.A., and Evans, D.A. (2013). Alzheimer disease
in the United States (2010-2050) estimated using the 2010 census. Neurology
80, 1778–1783.
Hejl, A.-M., Glenthøj, B., Mackeprang, T., Hemmingsen, R., and Waldemar, G.
(2004). Prepulse inhibition in patients with Alzheimer's disease. Neurobiol. Aging
25, 1045–1050.
Hellner, K., Walther, T., Schubert, M., and Albrecht, D. (2005). Angiotensin-(1-7)
enhances LTP in the hippocampus through the G-protein-coupled receptor Mas.
Mol. Cell. Neurosci. 29, 427–435.
Hemming, M.L., Selkoe, D.J., and Farris, W. (2007). Effects of prolonged
angiotensin-converting enzyme inhibitor treatment on amyloid β-protein
metabolism in mouse models of Alzheimer disease. Neurobiol. Dis. 26, 273–281.
Heneka, M.T., Reyes-Irisarri, E., Hüll, M., and Kummer, M.P. (2011). Impact and
Therapeutic Potential of PPARs in Alzheimer's Disease. Current
Neuropharmacology 9, 643–650.
Heringer-Walther, S., Eckert, K., Schumacher, S.-M., Uharek, L., Wulf-
Goldenberg, A., Gembardt, F., Fichtner, I., Schultheiss, H.-P., Rodgers, K., and
Walther, T. (2009). Angiotensin-(1-7) stimulates hematopoietic progenitor cells in
vitro and in vivo. Haematologica 94, 857–860.
Hickey, F., Hickey, E., and Summar, K.L. (2012). Medical update for children with
Down syndrome for the pediatrician and family practitioner. Adv Pediatr 59, 137–
157.
220
Hill, J.M., Zalos, G., Halcox, J.P.J., Schenke, W.H., Waclawiw, M.A., Quyyumi,
A.A., and Finkel, T. (2003). Circulating endothelial progenitor cells, vascular
function, and cardiovascular risk. N. Engl. J. Med. 348, 593–600.
Hock, C., Villringer, K., Müller-Spahn, F., Wenzel, R., Heekeren, H., Schuh-
Hofer, S., Hofmann, M., Minoshima, S., Schwaiger, M., Dirnagl, U., et al. (1997).
Decrease in parietal cerebral hemoglobin oxygenation during performance of a
verbal fluency task in patients with Alzheimer's disease monitored by means of
near-infrared spectroscopy (NIRS) — correlation with simultaneous rCBF-PET
measurements. Brain Res. 755, 293–303.
HOOPER, N.M., and Turner, A.J. (2003). An ACE structure. Nat. Struct. Biol. 10,
155–157.
Hossmann, K.A. (1994). Viability thresholds and the penumbra of focal ischemia.
Ann. Neurol.
Hristov, M., Erl, W., and Weber, P.C. (2003). Endothelial progenitor cells:
mobilization, differentiation, and homing. Arterioscler Thromb Vasc Biol 23,
1185–1189.
Hristov, M., Schmitz, S., Schuhmann, C., Leyendecker, T., Hundelshausen, von,
P., Krötz, F., Sohn, H.-Y., Nauwelaers, F.A., and Weber, C. (2009). An optimized
flow cytometry protocol for analysis of angiogenic monocytes and endothelial
progenitor cells in peripheral blood. Cytometry A 75, 848–853.
Huijing Xia, E.L. (2008). Angiotensin converting enzyme 2 in the brain: properties
and future directions. Journal of Neurochemistry 107, 1482–1494.
Hye, A., Riddoch-Contreras, J., Baird, A.L., Ashton, N.J., Bazenet, C., Leung, R.,
Westman, E., Simmons, A., Dobson, R., Sattlecker, M., et al. (2014). Plasma
proteins predict conversion to dementia from prodromal disease. Alzheimer's &
Dementia 10, 799–807.e2.
Iadecola, C. (2013). The pathobiology of vascular dementia. Neuron 80, 844–
866.
Iadecola, C., and Gorelick, P.B. (2004). Hypertension, Angiotensin, and Stroke:
Beyond Blood Pressure. Stroke 35, 348–350.
Ikhapoh, I.A., and Pelham, C.J. (2015). Synergistic effect of angiotensin II on
vascular endothelial growth factor-A-mediated differentiation of bone marrow-
derived mesenchymal stem cells into …. … Cell Research & ….
221
Imamura, H., Ohta, T., Tsunetoshi, K., Doi, K., Nozaki, K., Takagi, Y., and Kikuta,
K.-I. (2010). Transdifferentiation of bone marrow-derived endothelial progenitor
cells into the smooth muscle cell lineage mediated by tansforming growth factor-
beta1. Atherosclerosis 211, 114–121.
Imanishi, T., Hano, T., and Nishio, I. (2004). Angiotensin II potentiates vascular
endothelial growth factor-induced proliferation and network formation of
endothelial progenitor cells. Hypertension Research 27, 101–108.
Imanishi, T., Hano, T., and Nishio, I. (2005). Angiotensin II accelerates
endothelial progenitor cell senescence through induction of oxidative stress. J.
Hypertens. 23, 97–104.
Irizarry, M.C. (2004). Biomarkers of Alzheimer disease in plasma.
Neurotherapeutics 1, 226–234.
Iyer, S.N., Ferrario, C.M., and Chappell, M.C. (1998). Angiotensin-(1-7)
contributes to the antihypertensive effects of blockade of the renin-angiotensin
system. 31, 356–361.
Izzo, J.L., and Zion, A.S. (2011). Value of Angiotensin receptor blocker therapy in
diabetes. J Clin Hypertens (Greenwich) 13, 290–295.
Jack, C.R., Jr, Knopman, D.S., Jagust, W.J., Shaw, L.M., Aisen, P.S., Weiner,
M.W., Petersen, R.C., and Trojanowski, J.Q. (2010). Hypothetical model of
dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurology
9, 119–128.
Jagust, W.J. (2000). Neuroimaging in dementia. Neurol Clin 18, 885–902.
Jak, A.J., Bondi, M.W., Delano-Wood, L., Wierenga, C., Corey-Bloom, J.,
Salmon, D.P., and Delis, D.C. (2009). Quantification of five neuropsychological
approaches to defining mild cognitive impairment. Am J Geriatr Psychiatry 17,
368–375.
Janelsins, M.C., Kohli, S., Mohile, S.G., Usuki, K., Ahles, T.A., and Morrow, G.R.
(2011). An update on cancer- and chemotherapy-related cognitive dysfunction:
current status. Semin. Oncol. 38, 431–438.
Jankowsky, J.L., Younkin, L.H., Gonzales, V., Fadale, D.J., Slunt, H.H., Lester,
H.A., Younkin, S.G., and Borchelt, D.R. (2007). Rodent A beta modulates the
solubility and distribution of amyloid deposits in transgenic mice. J. Biol. Chem.
282, 22707–22720.
222
Ji, H., Atchison, L., Chen, Z., Chakraborty, S., Jung, Y., Truskey, G.A.,
Christoforou, N., and Leong, K.W. (2016). Transdifferentiation of human
endothelial progenitors into smooth muscle cells. Biomaterials 85, 180–194.
Jiang, Q., Lee, C.Y.D., Mandrekar, S., Wilkinson, B., Cramer, P., Zelcer, N.,
Mann, K., Lamb, B., Willson, T.M., Collins, J.L., et al. (2008). ApoE Promotes the
Proteolytic Degradation of Aβ. Neuron 58, 681–693.
Jiang, T., Tan, L., Gao, Q., Lu, H., Zhu, X.-C., Zhou, J.-S., and Zhang, Y.-D.
(2016). Plasma Angiotensin-(1-7) is a Potential Biomarker for Alzheimer's
Disease. Curr Neurovasc Res 13, 96–99.
Jiang, T., Zhang, Y.-D., Zhou, J.-S., Zhu, X.-C., Tian, Y.-Y., Zhao, H.-D., Lu, H.,
Gao, Q., Tan, L., and Yu, J.-T. (2015). Angiotensin-(1-7) is Reduced and
Inversely Correlates with Tau Hyperphosphorylation in Animal Models of
Alzheimer’s Disease. Mol Neurobiol.
Jonsson, T., Atwal, J.K., Steinberg, S., Snaedal, J., Jonsson, P.V., Bjornsson, S.,
Stefansson, H., Sulem, P., Gudbjartsson, D., Maloney, J., et al. (2012). A
mutation in APP protects against Alzheimer’s disease and age-related cognitive
decline. Nature 488, 96–99.
Jung, K.-H., Chu, K., Lee, S.-T., Song, E.-C., Sinn, D.-I., Kim, J.-M., Kim, S.-J.,
Kim, J.-H., Kang, K.-M., Park, H.-K., et al. (2008). Identification of neuronal
outgrowth cells from peripheral blood of stroke patients. Ann. Neurol. 63, 312–
322.
Kazama, K., Anrather, J., Zhou, P., Girouard, H., Frys, K., Milner, T.A., and
Iadecola, C. (2004). Angiotensin II Impairs Neurovascular Coupling in Neocortex
Through NADPH Oxidase–Derived Radicals. Circulation.
Kazama, K., Wang, G., Frys, K., Anrather, J., and Iadecola, C. (2003).
Angiotensin II attenuates functional hyperemia in the mouse somatosensory
cortex. Am. J. Physiol. Heart Circ. Physiol. 285, H1890–H1899.
Keene, C.D., Montine, T.J., and Kuller, L.H. (2016). Epidemiology, pathology,
and pathogenesis of Alzheimer disease (UpToDate).
Kehoe, P.G., Wong, S., AL Mulhim, N., Palmer, L.E., and Miners, J.S. (2016).
Angiotensin-converting enzyme 2 is reduced in Alzheimer's disease in
association with increasing amyloid-β and tau pathology. Alzheimers Res Ther 8,
50.
223
Keller, S.R. (2004). Role of the Insulin-Regulated Aminopeptidase IRAP in Insulin
Action and Diabetes. Biological and Pharmaceutical Bulletin 27, 761–764.
Klickstein, L.B., Kaempfer, C.E., and Wintroub, B.U. (1982). The granulocyte-
angiotensin system. Angiotensin I-converting activity of cathepsin G. J. Biol.
Chem. 257, 15042–15046.
Kong, X.-D., Zhang, Y., Liu, L., Sun, N., Zhang, M.-Y., and Zhang, J.-N. (2011).
Endothelial progenitor cells with Alzheimer's disease. Chin. Med. J. 124, 901–
906.
Krause, T., Lovibond, K., Caulfield, M., McCormack, T., and Williams, B. (2011).
Management of hypertension: summary of NICE guidance. Bmj 343, d4891–
d4891.
Krob, H.A., Vinsant, S.L., Ferrario, C.M., and Friedman, D.P. (1998).
Angiotensin-(1-7) immunoreactivity in the hypothalamus of the (mRen-2d)27
transgenic rat. Brain Res. 798, 36–45.
Kunz, J., Krause, D., Kremer, M., and Dermietzel, R. (1994). The 140-kDa
protein of blood-brain barrier-associated pericytes is identical to aminopeptidase
N. Journal of Neurochemistry 62, 2375–2386.
la Torre, de, J.C. (2004). Is Alzheimer's disease a neurodegenerative or a
vascular disorder? Data, dogma, and dialectics. The Lancet Neurology 3, 184–
190.
Larson, B. (2016a). Risk factors for cognitive decline and dementia (UpToDate).
Larson, E.B. (2016b). Evaluation of cognitive impairment and dementia
(UpToDate).
Launer, L.J., Ross, G.W., Petrovitch, H., Masaki, K., Foley, D., White, L.R., and
Havlik, R.J. (2000). Midlife blood pressure and dementia: the Honolulu-Asia
aging study. Neurobiol. Aging 21, 49–55.
Lee, S.-T., Chu, K., Jung, K.-H., Park, H.-K., Kim, D.-H., Bahn, J.-J., Kim, J.-H.,
Oh, M.-J., Lee, S.K., Kim, M., et al. (2009). Reduced circulating angiogenic cells
in Alzheimer disease. Neurology 72, 1858–1863.
Lee, S.-T., Chu, K., Jung, K.-H., Jeon, D., Bahn, J.-J., Kim, J.-H., Kun Lee, S.,
Kim, M., and Roh, J.-K. (2010). Dysfunctional characteristics of circulating
angiogenic cells in Alzheimer's disease. J. Alzheimers Dis. 19, 1231–1240.
224
Leissring, M.A., Farris, W., Chang, A.Y., Walsh, D.M., Wu, X., Sun, X., Frosch,
M.P., and Selkoe, D.J. (2003). Enhanced proteolysis of beta-amyloid in APP
transgenic mice prevents plaque formation, secondary pathology, and premature
death. Neuron 40, 1087–1093.
Li, N.-C., Lee, A., Whitmer, R.A., Kivipelto, M., Lawler, E., Kazis, L.E., and
Wolozin, B. (2010). Use of angiotensin receptor blockers and risk of dementia in
a predominantly male population: prospective cohort analysis. Bmj 340, b5465–
b5465.
Li, W., Peng, H., Seth, D.M., and Feng, Y. (2012). The Prorenin and (Pro)renin
Receptor: New Players in the Brain Renin-Angiotensin System? Int J Hypertens
2012, 1–8.
Lista, S., Faltraco, F., Prvulovic, D., and Hampel, H. (2013). Blood and plasma-
based proteomic biomarker research in Alzheimer's disease. Progress in
Neurobiology 101-102, 1–17.
Liu, S., Liu, J., Miura, Y., Tanabe, C., Maeda, T., Terayama, Y., Turner, A.J.,
Zou, K., and Komano, H. (2014). Conversion of Aβ43 to Aβ40 by the successive
action of angiotensin‐converting enzyme 2 and angiotensin‐converting enzyme.
J. Neurosci. Res. 92, 1178–1186.
Llorens-Cortes, C., and Mendelsohn, F.A.O. (2002). Organisation and functional
role of the brain angiotensin system. J Renin Angiotensin Aldosterone Syst 3
Suppl 1, S39–S48.
Loges, S., Fehse, B., and Brockmann, M.A. (2004). Identification of the adult
human hemangioblast. Stem Cells and … 13, 229–242.
Lomez, E.S.L., Araujo, R.C., Bader, M., Pesquero, J.B., and Pesquero, J.L.
(2002). Tonin and kallikrein in the brain of transgenic rat line expressing human
tissue kallikrein. 39, 229–232.
Luc, G., Arveiler, D., Evans, A., Amouyel, P., Ferrieres, J., Bard, J.-M., Elkhalil,
L., Fruchart, J.-C., Ducimetiere, P., PRIME Study Group (2003). Circulating
soluble adhesion molecules ICAM-1 and VCAM-1 and incident coronary heart
disease: the PRIME Study. Atherosclerosis 170, 169–176.
Luong, E., and Gerecht, S. (2009). Stem cells and scaffolds for vascularizing
engineered tissue constructs. Adv. Biochem. Eng. Biotechnol. 114, 129–172.
Machaliński, B., Paczkowska, E., Koziarska, D., and Ratajczak, M.Z. (2006).
225
Mobilization of human hematopoietic stem/progenitor-enriched CD34+ cells into
peripheral blood during stress related to ischemic stroke. Folia Histochem.
Cytobiol. 44, 97–101.
Madani, R., Poirier, R., Wolfer, D.P., Welzl, H., Groscurth, P., Lipp, H.-P., Lu, B.,
Mouedden, El, M., Mercken, M., Nitsch, R.M., et al. (2006). Lack of neprilysin
suffices to generate murine amyloid-like deposits in the brain and behavioral
deficit in vivo. J. Neurosci. Res. 84, 1871–1878.
Maler, J.M., Spitzer, P., Lewczuk, P., Kornhuber, J., Herrmann, M., and Wiltfang,
J. (2006). Decreased circulating CD34+ stem cells in early Alzheimer's disease:
evidence for a deficient hematopoietic brain support? Mol Psychiatry 11, 1113–
1115.
Mallat, S.G. (2012). What is a preferred angiotensin II receptor blocker-based
combination therapy for blood pressure control in hypertensive patients with
diabetic and non-diabetic renal impairment? Cardiovascular Diabetology 2012
11:1 11, 32.
Mancia, G., Laurent, S., Agabiti-Rosei, E., Ambrosioni, E., Burnier, M., Caulfield,
M.J., Cifkova, R., Clément, D., Coca, A., Dominiczak, A., et al. (2009).
Reappraisal of European guidelines on hypertension management: a European
Society of Hypertension Task Force document. J. Hypertens. 27, 2121–2158.
Mandrekar-Colucci, S., and Landreth, G.E. (2010). Microglia and inflammation in
Alzheimer's disease. CNS & Neurological Disorders Drug Targets 9, 156–167.
Mandrekar-Colucci, S., Karlo, J.C., and Landreth, G.E. (2012). Mechanisms
Underlying the Rapid Peroxisome Proliferator-Activated Receptor-γ-Mediated
Amyloid Clearance and Reversal of Cognitive Deficits in a Murine Model of
Alzheimer's Disease. The Journal of Neuroscience : the Official Journal of the
Society for Neuroscience 32, 10117–10128.
Mario, É.G., Santos, S.H.S., Ferreira, A.V.M., Bader, M., Santos, R.A.S., and
Botion, L.M. (2012). Angiotensin-(1-7) Mas-receptor deficiency decreases
peroxisome proliferator-activated receptor gamma expression in adipocytes. -
PubMed - NCBI. Peptides 33, 174–177.
Markowitsch, H.J., and Staniloiu, A. (2012). Amnesic disorders. Lancet 380,
1429–1440.
Maruyama, T., Uchida, K., and Hara, H. (1987). Suppressive effect of antibiotics
on colony formation from human megakaryocyte progenitors (CFU-M) and
226
granulocyte-macrophage progenitors (CFU-GM). Jpn. J. Pharmacol. 43, 423–
428.
Masuda, H., and Asahara, T. (2003). Post-natal endothelial progenitor cells for
neovascularization in tissue regeneration. Cardiovasc. Res. 58, 390–398.
Materson, B.J., Reda, D.J., and Cushman, W.C. (1995). Department of veterans
affairs single-drug therapy of hypertension studyRevised figures and new data.
Am. J. Hypertens. 8, 189–192.
Maurer, K., Volk, S., and Gerbaldo, H. (1997). Auguste D and Alzheimer's
disease. Lancet.
McDaniel, K.D., and Edland, S.D. (1995). Relationship Between Level of Insight
and Severity of Dementia in Alzheimer Disease. Alzheimer Disease & ….
McEvoy, L.K., Holland, D., Hagler, D.J., Fennema-Notestine, C., Brewer, J.B.,
Dale, A.M., Alzheimer's Disease Neuroimaging Initiative (2011). Mild cognitive
impairment: baseline and longitudinal structural MR imaging measures improve
predictive prognosis. Radiology 259, 834–843.
McKhann, G.M., Knopman, D.S., Chertkow, H., Hyman, B.T., Jack, C.R., Jr.,
Kawas, C.H., Klunk, W.E., Koroshetz, W.J., Manly, J.J., Mayeux, R., et al.
(2011). The diagnosis of dementia 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 7, 263–269.
McKinley, M.J., Albiston, A.L., Allen, A.M., Mathai, M.L., May, C.N., McAllen,
R.M., Oldfield, B.J., Mendelsohn, F.A.O., and Chai, S.Y. (2003). The brain renin-
angiotensin system: location and physiological roles. Int. J. Biochem. Cell Biol.
35, 901–918.
McPhee, S.J., and Hammer, G.D. (2009). Pathophysiology of Disease An
Introduction to Clinical Medicine, Sixth Edition (McGraw-Hill Medical).
Mead, L.E., Prater, D., Yoder, M.C., and Ingram, D.A. (2008). Isolation and
characterization of endothelial progenitor cells from human blood. Curr Protoc
Stem Cell Biol Chapter 2, Unit2C.1.
Medina, R.J., O'Neill, C.L., Sweeney, M., Guduric-Fuchs, J., Gardiner, T.A.,
Simpson, D.A., and Stitt, A.W. (2010). Molecular analysis of endothelial
progenitor cell (EPC) subtypes reveals two distinct cell populations with different
227
identities. BMC Med Genomics 3, 18.
Mentis, M.J., Alexander, G.E., Krasuski, J., Pietrini, P., Furey, M.L., Schapiro,
M.B., and Rapoport, S.I. (1998). Increasing required neural response to expose
abnormal brain function in mild versus moderate or severe Alzheimer's disease:
PET study using parametric visual stimulation. Am J Psychiatry 155, 785–794.
Michel, M.C., Foster, C., Brunner, H.R., and Liu, L. (2013). A systematic
comparison of the properties of clinically used angiotensin II type 1 receptor
antagonists. Pharmacol. Rev. 65, 809–848.
Miraglia, S., Godfrey, W., Yin, A.H., Atkins, K., Warnke, R., Holden, J.T., Bray,
R.A., Waller, E.K., and Buck, D.W. (1997). A novel five-transmembrane
hematopoietic stem cell antigen: isolation, characterization, and molecular
cloning. Blood 90, 5013–5021.
Mizrahi, R., Starkstein, S.E., Jorge, R., and Robinson, R.G. (2006).
Phenomenology and clinical correlates of delusions in Alzheimer disease. Am J
Geriatr Psychiatry 14, 573–581.
Mogi, M., and Horiuchi, M. (2009). Effects of angiotensin II receptor blockers on
dementia. Hypertens. Res. 32, 738–740.
Mogi, M., and Horiuchi, M. (2013). Effect of angiotensin II type 2 receptor on
stroke, cognitive impairment and neurodegenerative diseases. Geriatr Gerontol
Int 13, 13–18.
Mogi, M., Iwanami, J., and Horiuchi, M. (2012). Roles of Brain Angiotensin II in
Cognitive Function and Dementia. Int J Hypertens 2012, 1–7.
Mund, J.A., Estes, M.L., Yoder, M.C., Ingram, D.A., and Case, J. (2012). Flow
Cytometric Identification and Functional Characterization of Immature and
Mature Circulating Endothelial Cells. Arterioscler Thromb Vasc Biol 32, 1045–
1053.
Nguyen, M.T.X., Okina, E., Chai, X., Tan, K.H., Hovatta, O., Ghosh, S., and
Tryggvason, K. (2016). Differentiation of Human Embryonic Stem Cells to
Endothelial Progenitor Cells on Laminins in Defined and Xeno-free Systems.
Stem Cell Reports 7, 802–816.
Nishimura, Y., Ito, T., Hoe, K., and Saavedra, J.M. (2000). Chronic peripheral
administration of the angiotensin II AT(1) receptor antagonist candesartan blocks
brain AT(1) receptors. Brain Res. 871, 29–38.
228
Noble, F., Banisadr, G., Jardinaud, F., Popovici, T., Lai-Kuen, R., Chen, H.,
Bischoff, L., Parsadaniantz, S.M., Fournié-Zaluski, M.C., and Roques, B.P.
(2001). First discrete autoradiographic distribution of aminopeptidase N in
various structures of rat brain and spinal cord using the selective iodinated
inhibitor [125I]RB 129. Neuroscience 105, 479–488.
O'Caoimh, R., Kehoe, P.G., and Molloy, D.W. (2014). Renin Angiotensin
aldosterone system inhibition in controlling dementia-related cognitive decline. J.
Alzheimers Dis. 42, S575–S586.
Oddo, S., Caccamo, A., Shepherd, J.D., Murphy, M.P., Golde, T.E., Kayed, R.,
Metherate, R., Mattson, M.P., Akbari, Y., and LaFerla, F.M. (2003). Triple-
transgenic model of Alzheimer's disease with plaques and tangles: intracellular
Abeta and synaptic dysfunction. Neuron 39, 409–421.
Oppong, S.Y., and Hooper, N.M. (1993). Characterization of a secretase activity
which releases angiotensin-converting enzyme from the membrane. Biochemical
Journal 292, 597–603.
O’Bryant, S.E. (2010). A serum protein-based algorithm for the detection of
Alzheimer disease. - PubMed - NCBI. Arch Neurol 67, 1077.
Paczkowska, E., Larysz, B., Rzeuski, R., Karbicka, A., Jałowiński, R.,
Kornacewicz-Jach, Z., Ratajczak, M.Z., and Machaliński, B. (2005). Human
hematopoietic stem/progenitor-enriched CD34(+) cells are mobilized into
peripheral blood during stress related to ischemic stroke or acute myocardial
infarction. Eur. J. Haematol. 75, 461–467.
Palange, P., Testa, U., Huertas, A., Calabrò, L., Antonucci, R., Petrucci, E.,
Pelosi, E., Pasquini, L., Satta, A., Morici, G., et al. (2006). Circulating
haemopoietic and endothelial progenitor cells are decreased in COPD. Eur.
Respir. J. 27, 529–541.
Papinska, A.M., Mordwinkin, N.M., Meeks, C.J., Jadhav, S.S., and Rodgers, K.E.
(2015). Angiotensin-(1-7) administration benefits cardiac, renal, and progenitor
cell function in db/db mice. British Journal of Pharmacology n/a–n/a.
Park, T.S., and Zambidis, E.T. (2009). A role for the renin-angiotensin system in
hematopoiesis. Haematologica 94, 745–747.
Pelisch, N., Hosomi, N., Ueno, M., Masugata, H., Murao, K., Hitomi, H., Nakano,
D., Kobori, H., Nishiyama, A., and Kohno, M. (2010). Systemic candesartan
reduces brain angiotensin II via downregulation of brain renin-angiotensin
229
system. Hypertens. Res. 33, 161–164.
Pelliccia, F., Pasceri, V., Cianfrocca, C., Vitale, C., Speciale, G., Gaudio, C.,
Rosano, G.M.C., and Mercuro, G. (2010). Angiotensin II receptor antagonism
with telmisartan increases number of endothelial progenitor cells in normotensive
patients with coronary artery disease: a randomized, double-blind, placebo-
controlled study. Atherosclerosis 210, 510–515.
Pober, J.S. (2012). Just the FACS or Stalking the Elusive Circulating Endothelial
Progenitor Cell. Arterioscler Thromb Vasc Biol 32, 837–838.
Porto, I., Di Vito, L., De Maria, G.L., Dato, I., Tritarelli, A., Leone, A.M., Niccoli,
G., Capogrossi, M.C., Biasucci, L.M., and Crea, F. (2009). Comparison of the
effects of ramipril versus telmisartan on high-sensitivity C-reactive protein and
endothelial progenitor cells after acute coronary syndrome. Am. J. Cardiol. 103,
1500–1505.
Posner, H.B., Tang, M.-X., Luchsinger, J., Lantigua, R., Stern, Y., and Mayeux,
R. (2002). The relationship of hypertension in the elderly to AD, vascular
dementia, and cognitive function. Neurology 58, 1175–1181.
Powell, T.M. (2004). Granulocyte Colony-Stimulating Factor Mobilizes Functional
Endothelial Progenitor Cells in Patients With Coronary Artery Disease.
Arterioscler Thromb Vasc Biol 25, 296–301.
Prince, M.J. (2015). World Alzheimer Report 2015: the global impact of
dementia: an analysis of prevalence, incidence, cost and trends.
Prince, M., Bryce, R., Albanese, E., Wimo, A., Ribeiro, W., and Ferri, C.P.
(2013). The global prevalence of dementia: a systematic review and
metaanalysis. Alzheimers Dement 9, 63–75.e2.
Prohovnik, I., Mayeux, R., Sackeim, H.A., Smith, G., Stern, Y., and Alderson,
P.O. (1988). Cerebral perfusion as a diagnostic marker of early Alzheimer's
disease. Neurology 38, 931–931.
Prokopi, M., Pula, G., Mayr, U., Devue, C., Gallagher, J., Xiao, Q., Boulanger,
C.M., Westwood, N., Urbich, C., Willeit, J., et al. (2009). Proteomic analysis
reveals presence of platelet microparticles in endothelial progenitor cell cultures.
Blood 114, 723–732.
Rabelo, L.A., Alenina, N., and Bader, M. (2011). ACE2-angiotensin-(1-7)-Mas
axis and oxidative stress in cardiovascular disease. Hypertens. Res. 34, 154–
230
160.
Rafii, S., and Lyden, D. (2003). Therapeutic stem and progenitor cell
transplantation for organ vascularization and regeneration. Nat. Med. 9, 702–
712.
Ransohoff, R.M. (2016). A polarizing question: do M1 and M2 microglia exist?
Nat. Neurosci. 19, 987–991.
Ray, S., Britschgi, M., Herbert, C., Takeda-Uchimura, Y., Boxer, A., Blennow, K.,
Friedman, L.F., Galasko, D.R., Jutel, M., Karydas, A., et al. (2007). Classification
and prediction of clinical Alzheimer's diagnosis based on plasma signaling
proteins. Nat. Med. 13, 1359–1362.
Reinhard, H., Jacobsen, P.K., Lajer, M., Pedersen, N., Billestrup, N., Mandrup-
Poulsen, T., Parving, H.-H., and Rossing, P. (2010). Multifactorial treatment
increases endothelial progenitor cells in patients with type 2 diabetes.
Diabetologia 53, 2129–2133.
Relkin, N.R. (2015). Neuroimaging studies in the evaluation of dementia
(UpToDate).
Réaux, A., de Mota, N., Zini, S., Cadel, S., Fournié-Zaluski, M.C., Roques, B.P.,
Corvol, P., and Llorens-Cortès, C. (1999a). PC18, a specific aminopeptidase N
inhibitor, induces vasopressin release by increasing the half-life of brain
angiotensin III. Neuroendocrinology 69, 370–376.
Réaux, A., Fournié-Zaluski, M.C., David, C., Zini, S., Roques, B.P., Corvol, P.,
and Llorens-Cortès, C. (1999b). Aminopeptidase A inhibitors as potential central
antihypertensive agents. Proc. Natl. Acad. Sci. U.S.a. 96, 13415–13420.
RICE, G.I., THOMAS, D.A., GRANT, P.J., Turner, A.J., and HOOPER, N.M.
(2004). Evaluation of angiotensin-converting enzyme (ACE), its homologue
ACE2 and neprilysin in angiotensin peptide metabolism. Biochem. J. 383, 45–51.
Richardson, M.R., and Yoder, M.C. (2011). Endothelial progenitor cells: quo
vadis? Journal of Molecular and Cellular Cardiology 50, 266–272.
Rizos, C.V., Elisaf, M.S., Mikhailidis, D.P., and Liberopoulos, E.N. (2009). How
safe is the use of thiazolidinediones in clinical practice? Expert Opin Drug Saf 8,
15–32.
Robbins, M.E., Zhao, W., Garcia-Espinosa, M.A., and Diz, D.I. (2010). Renin-
angiotensin system blockers and modulation of radiation-induced brain injury.
231
Curr Drug Targets 11, 1413–1422.
Roberts, R.O., Knopman, D.S., Mielke, M.M., Cha, R.H., Pankratz, V.S.,
Christianson, T.J.H., Geda, Y.E., Boeve, B.F., Ivnik, R.J., Tangalos, E.G., et al.
(2014). Higher risk of progression to dementia in mild cognitive impairment cases
who revert to normal. Neurology 82, 317–325.
Rodgers, K.E., Espinoza, T., Roda, N., Meeks, C.J., Hill, C., Louie, S.G., and
Dizerega, G.S. (2012). Accelerated hematopoietic recovery with angiotensin-(1-
7) after total body radiation. Int. J. Radiat. Biol. 88, 466–476.
Rodgers, K.E., Oliver, J., and Dizerega, G.S. (2005). Phase I/II dose escalation
study of angiotensin 1-7 [A(1-7)] administered before and after chemotherapy in
patients with newly diagnosed breast cancer. Cancer Chemother. Pharmacol. 57,
559–568.
Rodgers, K.E., Xiong, S., and Dizerega, G.S. (2002). Accelerated recovery from
irradiation injury by angiotensin peptides. Cancer Chemother. Pharmacol. 49,
403–411.
Rodgers, K., Xiong, S., and Dizerega, G.S. (2003). Effect of angiotensin II and
angiotensin(1-7) on hematopoietic recovery after intravenous chemotherapy.
Cancer Chemother. Pharmacol. 51, 97–106.
Rogerson, F.M., Schlawe, I., Paxinos, G., Chai, S.Y., McKinley, M.J., and
Mendelsohn, F.A.O. (1995). Localization of angiotensin converting enzyme by in
vitro autoradiography in the rabbit brain. Journal of Chemical Neuroanatomy 8,
227–243.
Roks, A.J.M., Rodgers, K., and Walther, T. (2011). Effects of the renin
angiotensin system on vasculogenesis-related progenitor cells. Curr Opin
Pharmacol 11, 162–174.
Rowan, M.J., Klyubin, I., Cullen, W.K., and Anwyl, R. (2003). Synaptic plasticity
in animal models of early Alzheimer's disease. Philos. Trans. R. Soc. Lond., B,
Biol. Sci. 358, 821–828.
Royall, D.R., Espino, D.V., and Polk, M.J. (2004). Prevalence and patterns of
executive impairment in community dwelling Mexican Americans: results from the
Hispanic EPESE Study. … Journal of Geriatric ….
Rustemeyer, P., Wittkowski, W., Jurk, K., and Koller, A. (2006). Optimized flow
cytometric analysis of endothelial progenitor cells in peripheral blood. J
232
Immunoassay Immunochem 27, 77–88.
Saavedra, J.M. (2012). Angiotensin II AT(1) receptor blockers as treatments for
inflammatory brain disorders. Clin. Sci. 123, 567–590.
Saavedra, J.M., and Nishimura, Y. (1999). Angiotensin and Cerebral Blood Flow.
Cell. Mol. Neurobiol. 19, 553–573.
Safar, M.M., Arab, H.H., Rizk, S.M., and El-Maraghy, S.A. (2014). Bone Marrow-
Derived Endothelial Progenitor Cells Protect Against Scopolamine-Induced
Alzheimer-Like Pathological Aberrations. Mol Neurobiol 1–16.
Santos, C.R.A., Cardoso, I., and Gonçalves, I. (2011). Key Enzymes and
Proteins in Amyloid-Beta Production and Clearance.
Santos, R.A.S., Simoes e Silva, A.C., Maric, C., Silva, D.M.R., Machado, R.P.,
de Buhr, I., Heringer-Walther, S., Pinheiro, S.V.B., Lopes, M.T., Bader, M., et al.
(2003). Angiotensin-(1-7) is an endogenous ligand for the G protein-coupled
receptor Mas. Proc. Natl. Acad. Sci. U.S.a. 100, 8258–8263.
Sastre, M., Dewachter, I., Rossner, S., Bogdanovic, N., Rosen, E., Borghgraef,
P., Evert, B.O., Dumitrescu-Ozimek, L., Thal, D.R., Landreth, G., et al. (2006).
Nonsteroidal anti-inflammatory drugs repress beta-secretase gene promoter
activity by the activation of PPARgamma. Proc. Natl. Acad. Sci. U.S.a. 103, 443–
448.
Schinke, M., Baltatu, O., Böhm, M., Peters, J., Rascher, W., Bricca, G., Lippoldt,
A., Ganten, D., and Bader, M. (1999). Blood pressure reduction and diabetes
insipidus in transgenic rats deficient in brain angiotensinogen. Proc. Natl. Acad.
Sci. U.S.a. 96, 3975–3980.
Schroeter, M.L., Stein, T., Maslowski, N., and Neumann, J. (2009). Neural
correlates of Alzheimer's disease and mild cognitive impairment: a systematic
and quantitative meta-analysis involving 1351 patients. NeuroImage 47, 1196–
1206.
Segal, M.S., Shah, R., Afzal, A., Perrault, C.M., Chang, K., Schuler, A., Beem,
E., Shaw, L.C., Calzi, S.L., Harrison, J.K., et al. (2006). Nitric Oxide
Cytoskeletal–Induced Alterations Reverse the Endothelial Progenitor Cell
Migratory Defect Associated With Diabetes. Diabetes 55, 102–109.
Shaffer, R.G., Greene, S., Arshi, A., Supple, G., Bantly, A., Moore, J.S., and
Mohler, E.R. (2006). Flow cytometric measurement of circulating endothelial
233
cells: the effect of age and peripheral arterial disease on baseline levels of
mature and progenitor populations. Cytometry B Clin Cytom 70, 56–62.
Shin, J.-W., Lee, J.K., Lee, J.E., Min, W.-K., Schuchman, E.H., Jin, H.K., and
Bae, J.-S. (2011). Combined Effects of Hematopoietic Progenitor Cell
Mobilization from Bone Marrow by Granulocyte Colony Stimulating Factor and
AMD3100 and Chemotaxis into the Brain Using Stromal Cell-Derived Factor-1α
in an Alzheimer's Disease Mouse Model. Stem Cells 29, 1075–1089.
Sieff, C.A. (2016). Introduction to recombinant hematopoietic growth factors
(UpToDate).
SIMA, A.A.F. (2010). Encephalopathies: the emerging diabetic complications.
Acta Diabetol 47, 279–293.
Sink, K.M., Leng, X., Williamson, J., Kritchevsky, S.B., Yaffe, K., Kuller, L.,
Yasar, S., Atkinson, H., Robbins, M., Psaty, B., et al. (2009). Angiotensin-
Converting Enzyme Inhibitors and Cognitive Decline in Older Adults With
Hypertension. Arch Intern Med 169, 1195.
Skoog, I., Lernfelt, B., Landahl, S., Palmertz, B., Andreasson, L.A., Nilsson, L.,
Persson, G., Odén, A., and Svanborg, A. (1996). 15-year longitudinal study of
blood pressure and dementia. Lancet 347, 1141–1145.
Snowdon, D.A., Nun Study (2003). Healthy aging and dementia: findings from
the Nun Study. Ann. Intern. Med. 139, 450–454.
Sopova, K., Gatsiou, K., Stellos, K., and Laske, C. (2014). Dysregulation of
neurotrophic and haematopoietic growth factors in Alzheimer's disease: from
pathophysiology to novel treatment strategies. Curr Alzheimer Res 11, 27–39.
Spalding, K.L., Bergmann, O., Alkass, K., Bernard, S., Salehpour, M., Huttner,
H.B., Boström, E., Westerlund, I., Vial, C., Buchholz, B.A., et al. (2013).
Dynamics of Hippocampal Neurogenesis in Adult Humans. Cell 153, 1219–1227.
Spandidos, A., Wang, X., Wang, H., and Seed, B. (2009). PrimerBank: a
resource of human and mouse PCR primer pairs for gene expression detection
and quantification. Nucleic Acids Research 38, D792–D799.
Spulber, G., Niskanen, E., Macdonald, S., Kivipelto, M., Padilla, D.F., Julkunen,
V., Hallikainen, M., Vanninen, R., Wahlund, L.-O., and Soininen, H. (2012).
Evolution of global and local grey matter atrophy on serial MRI scans during the
progression from MCI to AD. Curr Alzheimer Res 9, 516–524.
234
Stellos, K., Panagiota, V., Sachsenmaier, S., Trunk, T., Straten, G., Leyhe, T.,
Seizer, P., Geisler, T., Gawaz, M., and Laske, C. (2010). Increased Circulating
Progenitor Cells in Alzheimer's Disease Patients with Moderate to Severe
Dementia: Evidence for Vascular Repair and Tissue Regeneration? Journal of
Alzheimer's Disease 19, 591–600.
Stover, K.R., Campbell, M.A., and Van Winssen, C.M. (2015). Early detection of
cognitive deficits in the 3xTg-AD mouse model of Alzheimer's disease.
Behavioural Brain ….
Swerdlow, N.R., Weber, M., Qu, Y., Light, G.A., and Braff, D.L. (2008). Realistic
expectations of prepulse inhibition in translational models for schizophrenia
research. Psychopharmacology (Berl.) 199, 331–388.
Taguchi, A., Nakagomi, N., Matsuyama, T., Kikuchi-Taura, A., Yoshikawa, H.,
Kasahara, Y., Hirose, H., Moriwaki, H., Nakagomi, T., Soma, T., et al. (2008).
Circulating CD34-Positive Cells Have Prognostic Value for Neurologic Function
in Patients with past Cerebral Infarction. Journal of Cerebral Blood Flow &
Metabolism 29, 34–38.
Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., and
Yamanaka, S. (2007). Induction of pluripotent stem cells from adult human
fibroblasts by defined factors. Cell 131, 861–872.
Thum, T., Tsikas, D., Stein, S., Schultheiss, M., Eigenthaler, M., Anker, S.D.,
Poole-Wilson, P.A., Ertl, G., and Bauersachs, J. (2005). Suppression of
endothelial progenitor cells in human coronary artery disease by the endogenous
nitric oxide synthase inhibitor asymmetric dimethylarginine. J. Am. Coll. Cardiol.
46, 1693–1701.
Tota, S., Goel, R., Pachauri, S.D., N, R., Najmi, A.K., Hanif, K., and Nath, C.
(2013). Effect of angiotensin II on spatial memory, cerebral blood flow,
cholinergic neurotransmission, and brain derived neurotrophic factor in rats.
Psychopharmacology (Berl.) 226, 357–369.
Tota, S., Hanif, K., Kamat, P.K., Najmi, A.K., and Nath, C. (2012). Role of central
angiotensin receptors in scopolamine-induced impairment in memory, cerebral
blood flow, and cholinergic function. Psychopharmacology (Berl.) 222, 185–202.
Townamchai, N., Praditpornsilpa, K., and Eiam-Ong, S. (2010). Endothelial
progenitor cells in Asian kidney transplant patients. Transplant. Proc. 42, 1690–
1694.
235
Trojanowski, J.Q., Vandeerstichele, H., Korecka, M., Clark, C.M., Aisen, P.S.,
Petersen, R.C., Blennow, K., Soares, H., Simon, A., Lewczuk, P., et al. (2010).
Update on the biomarker core of the Alzheimer's Disease Neuroimaging Initiative
subjects. Alzheimer's & Dementia 6, 230–238.
Uekawa, K., Hasegawa, Y., Senju, S., Nakagata, N., Ma, M., Nakagawa, T.,
Koibuchi, N., and Kim-Mitsuyama, S. (2016). Intracerebroventricular Infusion of
Angiotensin-(1-7) Ameliorates Cognitive Impairment and Memory Dysfunction in
a Mouse Model of Alzheimer's Disease. J. Alzheimers Dis. 53, 127–133.
UEKI, A., GOTO, K., SATO, N., ISO, H., and MORITA, Y. (2006). Prepulse
inhibition of acoustic startle response in mild cognitive impairment and mild
dementia of Alzheimer type. Psychiatry Clin. Neurosci. 60, 55–62.
Unger, R.E., Krump-Konvalinkova, V., Peters, K., and Kirkpatrick, C.J. (2002). In
vitro expression of the endothelial phenotype: comparative study of primary
isolated cells and cell lines, including the novel cell line HPMEC-ST1.6R.
Microvasc. Res. 64, 384–397.
Van Dam, D., and De Deyn, P.P. (2006). Drug discovery in dementia: the role of
rodent models. Nat Rev Drug Discov.
Vanderstichele, H., Van Kerschaver, E., Hesse, C., Davidsson, P., Buyse, M.A.,
Andreasen, N., Minthon, L., Wallin, A., Blennow, K., and Vanmechelen, E.
(2000). Standardization of measurement of beta-amyloid(1-42) in cerebrospinal
fluid and plasma. Amyloid 7, 245–258.
Vicario, A., Martinez, C.D., Baretto, D., Casale, A.D., and Nicolosi, L. (2005).
Hypertension and Cognitive Decline: Impact on Executive Function. J Clin
Hypertens (Greenwich) 7, 598–604.
Walther, T., MENRAD, A., ORZECHOWSKI, H.-D., SIEMEISTER, G., Paul, M.,
and SCHIRNER, M. (2003). Differential regulation of in vivo angiogenesis by
angiotensin II receptors. Faseb J. 17, 2061–2067.
Wang, B.-R., Shi, J.-Q., Zhang, Y.-D., Zhu, D.-L., and Shi, J.-P. (2011).
Angiotensin II does not directly affect Aβ secretion or β-/γ-secretase activity via
activation of angiotensin II type 1 receptor. Neurosci. Lett. 500, 103–107.
Wang, C.-H., Verma, S., Hsieh, I.-C., Chen, Y.-J., Kuo, L.-T., Yang, N.-I., Wang,
S.-Y., Wu, M.-Y., Hsu, C.-M., Cheng, C.-W., et al. (2006a). Enalapril increases
ischemia-induced endothelial progenitor cell mobilization through manipulation of
the CD26 system. Journal of Molecular and Cellular Cardiology 41, 34–43.
236
Wang, D.-S., Dickson, D.W., and Malter, J.S. (2006b). beta-Amyloid degradation
and Alzheimer's disease. J. Biomed. Biotechnol. 2006, 58406–58412.
Wang, H., He, J., Zhang, R., Zhu, S., Wang, J., Kong, L., Tan, Q., and Li, X.-M.
(2012). Sensorimotor gating and memory deficits in an APP/PS1 double
transgenic mouse model of Alzheimer's disease. Behavioural Brain Research
233, 237–243.
Wang, J., Ho, L., Chen, L., Zhao, Z., Zhao, W., Qian, X., Humala, N., Seror, I.,
Bartholomew, S., Rosendorff, C., et al. (2007). Valsartan lowers brain beta-
amyloid protein levels and improves spatial learning in a mouse model of
Alzheimer disease. J. Clin. Invest. 117, 3393–3402.
Wang, S., Wang, R., Chen, L., Bennett, D.A., Dickson, D.W., and Wang, D.-S.
(2010a). Expression and functional profiling of neprilysin, insulin-degrading
enzyme, and endothelin-converting enzyme in prospectively studied elderly and
Alzheimer's brain. Journal of Neurochemistry 115, 47–57.
Wang, Y., Qian, C., Roks, A.J.M., Westermann, D., Schumacher, S.-M., Escher,
F., Schoemaker, R.G., Reudelhuber, T.L., van Gilst, W.H., Schultheiss, H.-P., et
al. (2010b). Circulating rather than cardiac angiotensin-(1-7) stimulates
cardioprotection after myocardial infarction. Circ Heart Fail 3, 286–293.
Wayner, M.J., Armstrong, D.L., Phelix, C.F., and Wright, J.W. (2001).
Angiotensin IV enhances LTP in rat dentate gyrus in vivo. Peptides.
Weksler, M.E., Szabo, P., Relkin, N.R., Reidenberg, M.M., Weksler, B.B., and
Coppus, A.M.W. (2013). Alzheimer“s disease and Down”s syndrome: treating
two paths to dementia. Autoimmun Rev 12, 670–673.
Werner, N., Kosiol, S., Schiegl, T., Ahlers, P., Walenta, K., Link, A., Böhm, M.,
and Nickenig, G. (2005). Circulating endothelial progenitor cells and
cardiovascular outcomes. N. Engl. J. Med. 353, 999–1007.
Wildsmith, K.R., Holley, M., Savage, J.C., Skerrett, R., and Landreth, G.E.
(2013). Evidence for impaired amyloid β clearance in Alzheimer's disease.
Alzheimers Res Ther 5, 33.
Wolk, D.A., and Dickerson, B.C. (2017). Clinical features and diagnosis of
Alzheimer disease (UpToDate).
Wright, J.W., and Harding, J.W. (2011). Brain renin-angiotensin—A new look at
an old system. Progress in Neurobiology 95, 49–67.
237
Wright, J.W., and Harding, J.W. (2013). The brain renin-angiotensin system: a
diversity of functions and implications for CNS diseases. Pflugers Arch. 465,
133–151.
Wright, J.W., Kawas, L.H., and Harding, J.W. (2013). A Role for the Brain RAS in
Alzheimer“s and Parkinson”s Diseases. Front Endocrinol (Lausanne) 4, 158.
Xu, P., Sriramula, S., and Lazartigues, E. (2011). ACE2/ANG-(1–7)/Mas pathway
in the brain: the axis of good. American Journal of ….
Yao, E.-H., Fukuda, N., Matsumoto, T., Kobayashi, N., Katakawa, M.,
Yamamoto, C., Tsunemi, A., Suzuki, R., Ueno, T., and Matsumoto, K. (2007).
Losartan improves the impaired function of endothelial progenitor cells in
hypertension via an antioxidant effect. Hypertens. Res. 30, 1119–1128.
Yin, A.H., Miraglia, S., Zanjani, E.D., Almeida-Porada, G., Ogawa, M., Leary,
A.G., Olweus, J., Kearney, J., and Buck, D.W. (1997). AC133, a novel marker for
human hematopoietic stem and progenitor cells. Blood 90, 5002–5012.
Yin, C., Li, S., Zhao, W., and Feng, J. (2013). Brain imaging of mild cognitive
impairment and Alzheimer's disease. Neural Regen Res 8, 435–444.
Yoder, M.C., Mead, L.E., Prater, D., Krier, T.R., Mroueh, K.N., Li, F., Krasich, R.,
Temm, C.J., Prchal, J.T., and Ingram, D.A. (2007). Redefining endothelial
progenitor cells via clonal analysis and hematopoietic stem/progenitor cell
principals. Blood 109, 1801–1809.
Yoshida, Y., Fukuda, N., Maeshima, A., Yamamoto, C., Matsumoto, T., Ueno, T.,
Nojima, Y., Matsumoto, K., and Soma, M. (2011). Treatment with valsartan
stimulates endothelial progenitor cells and renal label-retaining cells in
hypertensive rats. J. Hypertens. 29, 91–101.
Yoshitake, T., Kiyohara, Y., Kato, I., Ohmura, T., Iwamoto, H., Nakayama, K.,
Ohmori, S., Nomiyama, K., Kawano, H., Ueda, K., et al. (1995). Incidence and
risk factors of vascular dementia and Alzheimer's disease in a defined elderly
Japanese population: The Hisayama Study. Neurology 45, 1161–1168.
You, D., Cochain, C., Loinard, C., Vilar, J., Mees, B., Duriez, M., Lévy, B.I., and
Silvestre, J.-S. (2008). Hypertension impairs postnatal vasculogenesis: role of
antihypertensive agents. 51, 1537–1544.
Yöder, M.C. (2012). Human Endothelial Progenitor Cells. Cold Spring Harb
Perspect Med 2, a006692–a006692.
238
Yu, Y., Fukuda, N., Yao, E.-H., Matsumoto, T., Kobayashi, N., Suzuki, R., Tahira,
Y., Ueno, T., and Matsumoto, K. (2008). Effects of an ARB on endothelial
progenitor cell function and cardiovascular oxidation in hypertension. Am. J.
Hypertens. 21, 72–77.
Zambidis, E.T., Park, T.S., Yu, W., Tam, A., Levine, M., Yuan, X., Pryzhkova, M.,
and Péault, B. (2008). Expression of angiotensin-converting enzyme (CD143)
identifies and regulates primitive hemangioblasts derived from human pluripotent
stem cells. Blood 112, 3601–3614.
Zambidis, E.T., Sinka, L., Tavian, M., Jokubaitis, V., Park, T.S., Simmons, P.,
and Péault, B. (2007). Emergence of human angiohematopoietic cells in normal
development and from cultured embryonic stem cells. Ann. N. Y. Acad. Sci.
1106, 223–232.
Zhang, D., Xiao, Q., Luo, H., and Zhao, K. (2015). [Effects of angiotensin-(1-7)
on hippocampal expressions of GFAP and GDNF and cognitive function in rats
with diabetes mellitus]. Nan Fang Yi Ke Da Xue Xue Bao= ….
Zhu, D., Shi, J., Zhang, Y., Wang, B., Liu, W., Chen, Z., and Tong, Q. (2011).
Central angiotensin II stimulation promotes β amyloid production in Sprague
Dawley rats. PLoS ONE 6, e16037.
Zini, S., Fournié-Zaluski, M.C., Chauvel, E., Roques, B.P., Corvol, P., and
Llorens-Cortès, C. (1996). Identification of metabolic pathways of brain
angiotensin II and III using specific aminopeptidase inhibitors: predominant role
of angiotensin III in the control of vasopressin release. Proc. Natl. Acad. Sci.
U.S.a. 93, 11968–11973.
Zlokovic, B.V. (2005). Neurovascular mechanisms of Alzheimer's
neurodegeneration. Trends in Neurosciences 28, 202–208.

Abstract (if available)

Abstract Introduction: Alzheimer's disease (AD) is an insidious neurodegenerative disease that initially manifests as memory loss. Cognitive function progressively worsens over time, and ultimately results in death. Modifiable risk factors such as midlife hypertension can increase the risk of developing AD later in life. Clinical studies have found that angiotensin II receptor blockers (ARBs) may confer additional protection against cognitive decline beyond its antihypertensive actions. The mechanisms underlying this additional protection may lie beyond controlling blood pressure, but rather through the metabolism of angiotensin II to angiotensin-(1–7) [Ang-(1–7)]. Patients with AD also have reduced numbers of functional circulating angiogenic cells, perhaps reflecting diminished progenitor reserves. However, the role of these angiogenic cells early in cognitive decline is unknown. This dissertation consists of two main parts: (1) a preclinical study investigating the efficacy of the ARB candesartan and Ang-(1–7) in an animal model of AD and (2) an observational clinical study of circulating angiogenic cells in older adults with mild cognitive impairment (MCI). ❧ Methods: In the preclinical study, triple transgenic AD (3xTg-AD) mice were treated with candesartan, Ang-(1–7), or both for 8 months. Hippocampal amyloid-beta (Aβ) was assessed by immunoassay, and potential underlying mechanisms were investigated including: endothelial progenitor cells (EPCs), carotid blood flow, and Aβ clearance pathways. In the clinical study, adults (55 years of age or older) underwent a neuropsychological exam, MRI brain scan, and blood draw. Circulating angiogenic cells were characterized by flow cytometry and colony-forming assay. ❧ Results: 3xTg-AD mice treated with candesartan and Ang-(1–7) showed significant reductions in hippocampal insoluble Aβ. Candesartan increased circulating EPCs, reduced carotid blood flow, and increased the mRNA expression of Aβ-degrading enzymes insulin-degrading enzyme (IDE) and endothelin-converting enzyme 2 (ECE2), and microglial activation markers CD45 and Iba1. Participants with MCI exhibited significantly reduced progenitor cell subsets CD34 and CD133. Angiogenic cell subsets positively correlated with performance on verbal and visual memory tests. Voxel-wise regression analyses revealed that participants with higher CD34 cells correlated to increased volumes in the lingual gyrus, posterior cingulate gyrus, and precuneus. ❧ Conclusion: Long-term treatment with candesartan and Ang-(1–7) significantly reduced hippocampal Aβ in a mouse model of AD, which may be attributed to increased Aβ clearance potentially through the increased expression of Aβ-degrading enzymes IDE and ECE2. Circulating progenitor cells were found to be decreased in MCI, and positively correlated to memory tests and brain regions involved in verbal and visual memory. These results suggest that circulating progenitor cells may have a protective role in the early stages of cognitive decline, which may serve as a useful diagnostic, biomarker, or point of intervention.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Renin-angiotensin system modulation for the prevention and treatment of metabolic dysfunction

PDF

An iPSC-based biomarker strategy to identify neuroregenerative responders to allopregnanolone

PDF

Modulation of the renin-angiotensin system (RAS) for the treatment of COVID-19

PDF

The peptide angiotensin-(1-7) as a novel treatment for complications induced by type 2 diabetes mellitus

PDF

Angiotensin (1-7): a novel treatment for diabetes-induced kidney and heart dysfunction

PDF

Alterations in renin-angiotensin system signaling in SARS-CoV2 infection

PDF

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

PDF

Perimenopausal transition increases blood brain permeability: implications for neurodegenerative diseases

PDF

Clinical, functional and therapeutic analysis of CD99 in acute myeloid leukemia

PDF

New views, inputs, and properties: a new look at the renin-angiotensin system

PDF

Neuroinflammation and ApoE4 genotype in at-risk female aging: implications for Alzheimer's disease

PDF

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

PDF

The protective role of progenitor cells: cell proliferation relates to cognition and white matter microstructure

PDF

Indexing cerebrovascular resistance in cognitive decline and Alzheimer's disease

PDF

Investigating the effects of T cell mediated anti-leukemia activity in FLT3-ITD positive acute myeloid leukemia

PDF

Therapeutic potential of rhesus theta (θ) defensin-1 in cystic fibrosis airway infection and inflammation

PDF

Factors that may impact drug disposition and metabolism

PDF

Dihydromyricetin improves social isolation anxiety induced synaptic loss and astrogliosis in mice hippocampus

PDF

Allopregnanolone promotes cholesterol and amyloid-beta clearance mechanisms

PDF

Laser induced modulation of ocular renin-angiotensin system for treatment of dry age-related macular degeneration

Asset Metadata

Creator Tan, Alick (author)

Core Title Averting dementia: renin-angiotensin system and angiogenic cells in cognitive decline

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Clinical and Experimental Therapeutics

Publication Date 01/03/2019

Defense Date 05/31/2017

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

Tag Alzheimer's disease,OAI-PMH Harvest,progenitor cells,renin-angiotensin system

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Rodgers, Kathleen E. (committee chair), Davies, Daryl (committee member), Nation, Daniel A. (committee member)

Creator Email alickt@gmail.com,alicktan@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-395371

Unique identifier UC11265443

Identifier etd-TanAlick-5479.pdf (filename),usctheses-c40-395371 (legacy record id)

Legacy Identifier etd-TanAlick-5479.pdf

Dmrecord 395371

Document Type Dissertation

Rights Tan, Alick

Type texts

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

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

Repository Name University of Southern California Digital Library

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