• No results found

Multiple amyloid binding sites in Alzheimer brain and their interaction with synaptic and inflammatory mechanisms

N/A
N/A
Protected

Academic year: 2023

Share "Multiple amyloid binding sites in Alzheimer brain and their interaction with synaptic and inflammatory mechanisms"

Copied!
63
0
0

Loading.... (view fulltext now)

Full text

(1)

Thesis for doctoral degree (Ph.D.) 2015

Multiple Amyloid Binding Sites in

Alzheimer Brain and Their Interaction with Synaptic and Inflammatory

Mechanisms

Ruiqing Ni

Thesis for doctoral degree (Ph.D.) 2015RAmyloid, Synaptic, Inflammatory interaction in Alzheimer disease

(2)

From Department of Neurobiology, Care Sciences and Society, Center for Alzheimer Research,

Division of Translational Alzheimer Neurobiology, Karolinska Institutet, Stockholm, Sweden

MULTIPLE AMYLOID BINDING SITES IN ALZHEIMER BRAIN AND THEIR INTERACTION

WITH SYNAPTIC AND INFLAMMATORY MECHANISMS

Ruiqing Ni

Stockholm 2015

(3)

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet. Printed by [AJ E print AB]

© Ruiqing Ni, 2015

ISBN 978-91-7549-873-7

(4)

Multiple amyloid binding sites in Alzheimer brain and their interaction with synaptic and inflammatory

mechanisms

THESIS FOR DOCTORAL DEGREE (Ph.D.) The thesis will be defended at Hörsalen Novum 4

th

floor, Huddinge Tuesday, March 31

st

2015 at 9.30 am

By

Ruiqing Ni

Principal Supervisor:

Prof. Agneta Nordberg, MD, PhD Karolinska Institutet

Department of NVS

Center for Alzheimer Research Division of Translational Alzheimer Neurobiology

Co-supervisor(s):

Prof. Per-Göran Gillberg, PhD Karolinska Institutet

Department of NVS

Center for Alzheimer Research Division of Translational Alzheimer Neurobiology

Forskare Amelia Marutle, PhD Karolinska Institutet

Department of NVS

Center for Alzheimer Research Division of Translational Alzheimer Neurobiology

Opponent:

Prof. Victor Villemagne, MD, PhD Melbourne University, Austin Health, The Florey Institute

Department of Medicine Examination Board:

Docent Stina Syvänen, PhD Uppsala University

Department of Public Health and Caring Sciences Division of Geriatrics; Molecular Geriatrics/

Rudbeck laboratory

Prof. Tormod Fladby, MD, PhD Oslo University

Department of Neurology

Division of Faculty, Akershus University Hospital Docent Susanne Frykman, PhD

Karolinska Institutet Department of NVS

Center for Alzheimer Research Division of Neurogeriatrics

(5)

To my dear parents 献给亲爱的父母

“Ask and it will be given to you; seek and you will find;

knock and the door will be opened to you.” (Matthew 7:7)

(6)

ABSTRACT

Alzheimer’s disease (AD) is the most common type of dementia. A series of pathophyslogical changes start many years prior to the emergence of clinical symptoms. The main aims of this thesis were to investigate fibrillar amyloid-β imaging tracers that bind to the AD brain and the relationships between amyloid pathology, inflammation, and synaptic changes in AD.

Amyloid-β plaque deposition is one of the pathological hallmarks of AD. We demonstrated that amyloid positron emission tomography (PET) tracers 3H-Pittsburgh compound B (PIB), BTA- 1, florbetaben, florbetapir and AZD2184 detect a similar high-affinity site and a varying low- affinity binding site on fibrillar amyloid-β in postmortem sporadic AD brain. Autosomal dominant AD showed an additional binding site with AZD2184 in the frontal cortex and higher 3H-PIB binding in the striatum than in sporadic AD. Amyloid tracer binding to fibrillar Aβ was influenced by resveratrol and AZD2184 showed the greatest changes. These findings suggest a multiple binding site model for amyloid tracers in the AD brain (Papers I, II).

Inflammation is recognized to play a crucial role in AD. Cross-sectional microPET imaging in APPswe transgenic AD mice showed increased 11C-deuterium-L-deprenyl PET binding (astrocytosis) at 6 months compared to age-matched wild-type mice, prior to the increase in 11C-AZD2184 PET retention (amyloid-β plaque deposition) that occurred at 18-24 months, suggesting that astrocytosis is an early event in comparison to amyloid-β plaque deposition. In vitro autoradiography and immunochemistry staining confirmed age-related increases in Aβ deposits and indicated a context-dependent astrocytosis in transgenic AD mice (Paper III).

Mild cognitive impairment is prodromal stage of AD. We found that the combination of measurement of parietal glucose metabolism using the neurodegeneration biomarker 18F- fluorodeoxyglucose PET with analysis of total tau levels in cerebrospinal fluid provided the best prediction of patients with mild cognitive impairment converting to AD (Paper IV).

Aβ assemblies bind to α7 nicotinic acetylcholine receptors (nAChRs) and form complexes in the AD brain. 3H-PIB measurements showed increased fibrillar Aβ levels in the presence of α7 nAChR agonists, suggesting a specific interaction between fibrillar amyloid-β and α7 nAChRs, and α7 nAChR drugs may influence on the fibrillar Aβ-α7 nAChR interaction (Paper V).

In conclusion, clinical amyloid tracers detect multiple binding sites on fibrillar amyloid-β in the AD brain. Amyloid-β interacts with astrocytosis and synaptic sites. A deeper understanding of the subtle difference of amyloid-β binding sites in brain could facilitate the development of amyloid-β tracers and drugs for AD.

(7)

LIST OF PUBLICATIONS

This thesis is based on the following papers:

Paper I

Ni R, Gillberg PG, Bergfors A, Marutle A, Nordberg A.

Amyloid tracers detect multiple binding sites in Alzheimer's disease brain tissue.

Brain. 2013;136:2217-27.

Paper II

Ni R, Gillberg PG, Viitanen M, Myllykangas L, Bogdanovic N, Långström B, Nordberg A.

Discrimination between clinical amyloid tracers and resveratrol in familial and sporadic Alzheimer disease.

Manuscript

Paper III

Rodriguez-Vieitez, E*, Ni R*, Gulyás B, Tóth M, Häggkvist J, Halldin C, Voytenko L, Marutle A, Nordberg A. *contributed equally

Astrocytosis precedes amyloid plaque deposition in Alzheimer APPswe transgenic mouse brain: a correlative positron emission tomography and in vitro imaging study.

Submitted manuscript under revision in Eur J Nucl Med and Mol Imaging

Paper IV

Choo IH, Ni R, Schöll M, Wall A, Almkvist O, Nordberg A.

Combination of (18)F-FDG PET and cerebrospinal fluid biomarkers as a better predictor of the progression to Alzheimer's disease in mild cognitive impairment patients.

J Alzheimers Dis. 2013;33(4):929-39.

Paper V

Ni R, Marutle A, Nordberg A.

Modulation of alpha7 nicotinic acetylcholine receptor and fibrillar amyloid-beta interactions in Alzheimer's disease brain.

J Alzheimers Dis. 2013;33(3):841-51.

(8)

CONTENTS

1 Introduction ... 6

1.1 Alzheimer’s disease ... 6

1.1.1 Neuropathology ... 6

1.1.2 Genetics and risk factors ... 11

1.2 Positron emission tomography ... 12

1.2.1 FDG PET for measurement of glucose metabolism ... 12

1.2.2 Amyloid tracers for amyloid plaque deposition ... 13

1.2.3 Neuroinflammation ... 15

1.2.4 Tau ... 16

1.3 Revised suggested diagnostic criteria ... 16

1.4 MicroPET ... 17

1.5 Time courses of pathological events in AD ... 19

1.6 Current and future therapeutics ... 20

2 Aims of the thesis ... 22

3 Methodology ... 23

3.1 Ethical considerations ... 23

3.2 Patients and model system ... 23

3.2.1 MCI patients (Paper IV) ... 23

3.2.2 Post-mortem human brain tissue (Papers I, II, V) ... 24

3.2.3 APPswe mice (Paper III) ... 24

3.3 PET and data analysis (Paper III, IV) ... 25

3.4 CSF measurements (Paper IV) ... 26

3.5 Neuropsychological assessment (paper IV) ... 27

3.6 In vitro receptor binding study (Papers I, II, III and V) ... 27

3.7 Immunochemistry (Paper III) ... 29

3.8 ELISA (Paper I) ... 29

3.9 Statistics ... 29

4 Results and discussion ... 30

4.1 Multiple amyloid tracer binding sites in AD brains ... 30

4.2 Regional amyloid distribution in AD brains ... 32

4.3 Interactions between phenols and amyloid tracer binding sites ... 33

4.4 Early astrocytosis in APPswe AD mice ... 33

4.5 Predictive model for MCI converting to AD ... 35

4.6 Interaction between fibrillar Aβ and nicotinic receptors ... 37

5 Concluding remarks ... 39

6 Acknowledgements ... 41

7 References ... 44

(9)

LIST OF ABBREVIATIONS

Aβ Amyloid-beta

Ach Acetylcholine

AChEI Acetylcholinesterase inhibitor

AD Alzheimer’s disease

ADAD Autosomal dominant Alzheimer’s disease ADNI Alzheimer's Disease Neuroimaging Initiative

APOE Apolipoprotein E

APP Amyloid precursor protein

APParc Arctic amyloid precursor protein mutation (E693G)

APPswe Swedish amyloid precursor protein mutation (K670N/M671L) BACE β-site APP-cleaving enzyme

BBB Blood brain barrier

Bmax Maximum number of binding sites BP(ND) Non-displaceable binding potential

CAA Cerebral amyloid angiopathy

CMRglc Cerebral metabolic rate of glucose

CSF Cerebrospinal fluid

11C-L-DED 11C- deuterium-L-deprenyl

DIAN Dominantly inherited Alzheimer Network

DSM-V Diagnostic and Statistical Manual of Mental Disorders, 5th edition

EOAD Early-onset Alzheimer’s disease ELISA Enzyme-linked immunosorbent assay

EMA European Medicines Agency

FDA US Food and Drug Administration

18F-FDG 18F-fluro-2-deoxyglucose GFAP Glial fibrillary acidic protein

GWAS Genome-wide association sequencing

IWG International Working Group

Ki Inhibition constant

mAChR Muscarinic acetylcholine receptor

(10)

MAO-B Monoamine oxidase B

MCI Mild cognitive impairment

MMSE Mini mental state examination

MRI Magnetic resonance imaging

nAChR Nicotinic acethylcholine receptor NFTs Neurofibrillary tangles

NIA-AA National Institute on Aging-Alzheimer’s Association

NINCSD-ADRDA National Institutet of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association

NMDAR N-methyl-D-aspartate receptor

NP Neuritic plaque

NSAID Non-steroidal anti-inflammatory drug PET Positron emission tomography

PIB Pittsburgh compound B

PS Presenilin

p-Tau Phosphorylated tau

ROI Region of interest

sAD Sporadic Alzheimer’s disease SRTM Simplified reference tissue model

SUV Standard uptake value

TSPO Translocator protein

t-Tau Total tau

wt Wild-type

(11)

1 INTRODUCTION

1.1 ALZHEIMER’S DISEASE

Dementia is a general term for a set of symptoms that include decline in memory and ability to perform daily activities. Alzheimer’s disease (AD) is the most common cause of dementia; it accounts for 50-70 % of all cases. Other common types of dementia include vascular dementia, Lewy body dementia, Frontotemporal dementia, and Parkinson’s disease with dementia. Neurodegenerative disorders are associated with the deposition of abnormal toxic proteins in the brain (Taylor et al., 2002; Ross and Poirier, 2004). These proteins can activate a cascade of biochemical changes, which precede the clinical symptoms by many years. The pathological hallmarks of AD are the abnormal accumulation of extracellular amyloid-beta (Aβ) plaques, intracellular neurofibrillary tangles (NFTs) containing hyperphosphorylated tau, inflammation, and synaptic and neuronal loss. Clinically, AD is characterized by the progressive loss of memory and cognitive functions, which gradually affects the daily life of patients. It is estimated that there were over 36 million patients living with AD in 2012 worldwide, and this number will increase to over 100 million by 2050, imposing a substantial burden on society (Wimo et al., 2013). Thus there is an urgent need for the development of early detection tools and effective treatment strategies for AD.

1.1.1 Neuropathology 1.1.1.1 Amyloid

Aβ is produced by proteolytic processing of amyloid precursor protein (APP) on neurons and glial cells. In the amyloidogenic pathway, APP is cleaved by the β-site APP-cleaving enzyme (BACE) and γ-secretase complex, producing Aβ of various lengths, such as Aβ38, 40, 42, and pyroglutamate Aβ, and the remaining APP intracellular domain. The γ-secretase is a transmembrane protein complex, containing presenilin (PS), nicastrin, anterior pharynx defective-1 and PS enhancer-2. Aβ42 is more hydrophobic and prone to self-aggregation than Aβ40. Aβ is then released, cleared into the cerebrospinal fluid (CSF) and later removed into the blood (O'Brien and Wong, 2011). An imbalance between the production and clearance of Aβ leads to its abnormal accumulation in the brain in varying aggregation states: oligomers, protofibrils, fibrils and amyloid plaques (Lesné et al., 2006; Haass and Selkoe, 2007; Lambert et al., 2007;

Shankar et al., 2008) (Figure 1). After the accumulation stage, the amount of senile plaques may reach a plateau, with a dynamic balance between deposition and resolution

(12)

(Hyman et al., 1993). It has been estimated that the amount of Aβ is approximately 1-3 µmol in the cortices of AD patients (Naslund et al., 1994).

Figure 1 Different forms of Amyloid-β aggregates interact with tau, inflammation processes, astrocytes, microglia, synapses, and neurotransmitters in Alzheimer’s disease.

Findings of mutations in the APP and PS genes in the 1990s lead to the amyloid cascade hypothesis, which states that Aβ plays a key role in the cascade of pathology, causing downstream events leading to synaptic degeneration and neuronal loss (Haass et al., 1992; Shoji et al., 1992; Hardy and Selkoe, 2002). However, recent negative results in AD patients who received agents targeted at Aβ in phase II and III trials have raised question in regards to the validity of the amyloid hypothesis (Karran and Hardy, 2014). Many unresolved questions remain to be explored; for example, what is the role of Aβ in the cause and development of AD, which form (s) of Aβ is most toxic, and how does Aβ interact with other pathological features (Benilova et al., 2012)? Aβ is found both intracellularly and extracellularly (LaFerla et al., 2007). Soluble Aβ oligomers are currently considered to be the most toxic form, causing synaptic dysfunction in animal models (Haass and Selkoe, 2007; Lesne et al., 2013).Cortical pyroglutamate Aβ has also been shown to correlate with cognitive decline in AD (Pivtoraiko et al., 2015).

Whether Aβ has a physiological role under normal non-pathological conditions remains unclear. Functional magnetic resonance imaging (MRI) studies in humans have

suggested a link between Aβ and synaptic activity and have shown that the cortical hubs revealed by intrinsic functional connectivity overlap with the amyloid

deposition region (Buckner et al., 2009). In the transgenic AD mouse model, synaptic activity regulates interstitial fluid Aβ levels in vivo (Cirrito et al., 2005). Another study

(13)

has reported that Aβ is released in an activity-independent fashion from rat brain nerve terminals (Lundgren et al., 2014).

Different forms and types of amyloid plaque have been identified; these include classic senile plaques, diffuse plaques, and cerebral amyloid angiopathy (CAA). Neuritic plaques stained by Bielschowski silver contain degenerating neuronal processes and tau paired helical filaments, and may also contain reactive astrocytes and microglia (Braak and Braak, 1991). The deposition of Aβ plaques in the brain follows a regional

sequence, from neocortical regions to regions that receive neuronal projections, and later to subcortical regions and the cerebellum (Thal et al., 2002).

1.1.1.2 Tau and Neurofibrillary tangles

Microtubule-associated tau protein is produced by alternative splicing form a single gene on chromosome 17. Six tau isoforms exist in the human brain. The AD brain contains 3-repeat and 4-repeat isoforms (Iqbal et al., 2010). Tau is located inside the neurons where it plays a role in regulating neurite out-growth, axonal transport and microtubule stability, and interacts with other proteins such as protein kinase Fyn under physiological conditions (Morris et al., 2011). Tau is abnormally hyperphosphorylated under the pathological conditions associated with AD, to forms tau filaments and tangles, leading to the disruption of microtubule stability and cell death (Iqbal et al., 2010; Spillantini and Goedert, 2013). Tangles start to accumulate in the entorhinal cortex in the AD brain (Braak and Braak, 1991), and subsequently spread from neuron to neuron from the entorhinal cortex to the hippocampus and neocortex (Holmes et al., 2014). Whether tau aggregation is associated with a loss or gain of function and also which form of tau (soluble or misfolded) is more toxic are currently under debate (Morris et al., 2011). Experimental evidence has suggested that Aβ-induced tau pathology mediates toxicity in the dendrite, and shows a possible synergistic effect at the synapse resulting in dysfunction in cellular and animal models (Ittner and Gotz, 2011; Spires-Jones and Hyman, 2014).

1.1.1.3 Neuroinflammation

Inflammation is a homeostatic response that promotes the survival and function of the organism by detecting and removing pathogenic insults. Inflammation is a major component of many neurodegenerative diseases, including amyotrophic lateral sclerosis, Parkinson’s disease, multiple sclerosis and AD (Heneka et al., 2015).

(14)

Although long viewed as a secondary response to Aβ deposition and neuronal cell death in AD, increasing evidence indicates that neuroinflammation plays a primary role in the pathogenesis of AD (Akiyama et al., 2000). In post-mortem AD brains, reactive astrocytes and activated microglia are present in the vicinity of Aβ plaques (Figure 2).

Figure 2 Triple immunohistochemical staining of glial fibrillary acidic protein (GFAP+) reactive astrocytes (brown) co-localized with Aβ aggregates (red) and α7 nicotinic

acetylcholine receptor (nAChR) (blue) in the frontal cortex (A, B) and hippocampus (C) from an AD patient (Marutle et al., 2013).

Microglia and astrocytes are major players in brain inflammation. Activated microglia has been categorized as binarily into “classical activation, proinflammation” M1state or

“alternative activation, anti-inflammation” M2 state based on the expression or lack of expression of specific genes, although the pure M1, M2 state do not exist in reality.

Microglia are morphologically dynamic; they undergo priming according to the local context, and play a phagocytic role in the clearance of Aβ deposits (Perry and Holmes, 2014). Astrocytes account for 80 % of all cells in the human brain; different subtypes have been identified. Astrocytes are also important in maintaining brain homeostasis, neurogenesis, and the micro-architecture of the grey matter. A role for astrocytes in the clearance of Aβ deposits in AD has been suggested, as has altered astrocyte function in AD. Recent evidence suggests that the early stages of the

neurodegenerative process are associated with atrophy of astroglia, resulting in disruptions in synaptic connectivity, imbalance in neurotransmitter homeostasis, and neuronal death through increased excitotoxicity. In the later stages, astrocytes become activated and contribute to the neuroinflammatory component of neurodegeneration (Verkhratsky et al., 2010; 2014).

(15)

Altered microglia-astrocytes-neuron communication, and increased levels of inflammatory factors, complements, cytokines and chemokines are also observed (Glass et al., 2010). The chronic inflammation in AD can be considered as a

dysfunction in the braking mechanism forming a vicious circle with Aβ. Inflammatory regulators (such as NF-κB) promote amyloidogenic processing of APP; Aβ induces inflammation and activated glia; neuronal kinase activated by inflammation also promotes tau phosphorylation. In addition, neurovascular changes and blood brain barrier (BBB) leakage are observed in the AD brains (Zlokovic, 2011; Montagne et al., 2015). Whether neuroinflammation plays a beneficial or detrimental role, in relation to Aβ deposition requires further investigation (Wyss-Coray, 2006; Krstic and Knuesel, 2012). Recent evidence indicates that anti-inflammatory interleukin (IL)-10 has an unexpectedly negative effect on Aβ accumulation and on cognition in APP mouse models, suggesting a complex relationship between amyloid and inflammation (Chakrabarty et al., 2015).Systematic infections and inflammation may also drive the progression of AD (Perry et al., 2007).

1.1.1.4 Synaptic and neuronal degeneration

Disruption of synaptic function, neurotransmitter changes, synaptic loss and later neuronal loss are regarded as main pathological hallmarks of AD, correlating with memory deficits (Selkoe, 2002). The cholinergic system is impaired early in AD.

Cholinergic neurotransmission is mediated by the neurotransmitter acetylcholine (ACh) which interacts with neuronal nicotinic acetylcholine receptors (nAChRs) and muscarinic acetylcholine receptors (mAChRs). The nAChRs are widely distributed in the brain and play an important role in cognitive function, regulating

neurotransmission especially the α4β2 and α7 nAChR subtypes. Substantial

reductions in the number of nAChRs, and in ACh concentrations, and increases in the levels of acetylcholinesterase (AChE), have been reported in AD (Paterson and Nordberg, 2000). Nicotinic receptors are therefore a therapeutic target for AD, providing neuroprotective benefits (Dineley et al., 2015).

Aβ binds to synaptic sites, and impairs synaptic plasticity in the hippocampus.

The internalization/down-regulation of metabotropic glutamate receptors reduces excitatory spines and leads to neural network changes (Snyder et al., 2005). Various cell surface receptors including α7 nAChRs, N-methyl-D-aspartate receptors

(16)

(NMDARs), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs), and insulin receptors mediate the Aβ effects on synapses (Mucke and Selkoe, 2012) and it is important to identify these molecular targets to find therapeutic interventions for AD. Oligomeric and fibrillar Aβ are reported to bind with high affinity to α7 nAChRs, and Aβ may accumulate intracellularly through forming an Aβ- α7 nAChR complex (Wang et al., 2000; Buckingham et al., 2009; Lilja et al.,

2011).This complex may facilitate the internalization of Aβ, leading to intrasynaptic Aβ accumulation (Snyder et al., 2005). Immunochemical staining in post-mortem AD brain demonstrated α7 nAChRs on Aβpositive cortical pyramidal neurons and

astrocytes, where an Aβ/α7 nAChR complex has also been detected inside Aβ plaques (Nagele et al., 2002; Yu et al., 2005).

1.1.2 Genetics and risk factors Autosomal dominant Alzheimer’s disease

Genetically, AD can be divided into autosomal dominant Alzheimer’s disease (ADAD) which makes up 1 % of AD cases, and sporadic AD (sAD). ADAD is caused by

mutations in the genes for PS1 on chromosome 14, PS2 on chromosome 1 and APP gene on chromosome 21 (Bertram et al., 2010; Bettens et al., 2013). Among the more than 200 mutations that have been indentified are the Arctic APP mutation (APParc) (Basun et al., 2008), the Swedish APP mutation (APPswe KM 670/671NL) (Mullan et al., 1992), and the PS1 M146V mutation (Haltia et al., 1994). The location of the mutation on the gene influences the Aβ production and clearance rates, the Aβ42/ Aβ40 ratio, the aggregation rate, and the composition of the amyloid plaques (Karran et al., 2011; Bateman et al., 2012). APParc mutation carriers have increased protofibrillar Aβ that accumulates more intracellularly, forming ring-shaped plaques (Nilsberth et al., 2001; Basun et al., 2008; Schöll et al., 2012), while APPswe mutation carriers show levels of cortical Aβ similar to those in sporadic AD (Bogdanovic et al., 2002). PS1 M146V mutation carriers have more Aβ plaques and neurofibrillary tangles in the cortical and subcortical regions where PS1 exon 9 deletion (PS1 E∆9) mutation carriers have characteristic large ‘cotton wool’ plaques in the subcortical and cortical regions (Crook et al., 1998; Verkkoniemi et al., 2001). Studies in ADAD have been extremely important to our understanding of the biology and time course of AD.

(17)

Risk factors in sporadic AD

The etiology of AD is still not clear. Aging is the main risk factor, the prevalence of AD doubles every 5 years from 1 % at 65 years of age up to 50 % at 85 years of age (Ballard et al., 2011). The ε4 variant of Apolipoprotein E (APOE) has been identified as a major risk factor for sporadic AD (Liu et al., 2013). Recent sequencing and pathway analyses of genome-wide association sequencing (GWAS) identified several risk genes involved in three biological pathways (cholesterol metabolism, membrane endocytosis and the immune response pathway), including genes for CLU, PICAM, TREM2 and CR1 (Harold et al., 2009; Hollingworth et al., 2011; Naj et al., 2011;

Cruchaga et al., 2014). These findings suggest critical roles for different pathological processes in AD (Bettens et al., 2013; Karch and Goate, 2015). Epidemiological studies have estimated that one-third of AD cases might be attributable to potentially

modifiable risk factors, and the incidence of AD might be reduced with targeted intervention. These risk factors include vascular risk factors physical inactivity, smoking, midlife hypertension, obesity, diabetes and depression (Exalto et al., 2014;

Norton et al., 2014).

1.2 POSITRON EMISSION TOMOGRAPHY

Recent molecular imaging studies have provided valuable insights into the time course of AD pathology, such as the accumulation of amyloid plaque, NFTs, inflammation and synaptic changes. There are several prerequisites for an ideal positron emission tomography (PET) tracer; it should be selective, specific binding to the target with high affinity, be moderately lipophilic, crosses the BBB, be rapidly cleared from the brain, and leave no radiolabeled metabolites in the brain (Shah and Catafau, 2014). 11C- labeled tracers have a 20 min half life and 18F-labeled tracers 120 min half life. Thus the use of 11C tracers is limited to research hospitals that are equipped with an on-site cyclotron while 18F-labeled tracers could be transported to nearby hospitals, and more suitable for clinical use.

1.2.1 FDG PET for measurement of glucose metabolism

Glucose metabolism in the brain, a marker for neurodegeneration in AD, is measured by 18F-fluro-2-deoxyglucose (FDG) PET. Reductions in the cerebral metabolic rate for glucose consumption (CMRglc) have been observed in patients with mild cognitive

(18)

impairment (MCI), mainly in the posterior cingulate, parietal and temporal cortices compared to cognitively normal controls (Herholz et al., 2002; Drzezga et al., 2003;

Mosconi, 2005). The use of 18F-FDG-PET has allowed differentiation between and detection of different dementia disorders (Chetelat et al., 2003; Landau et al., 2010).

1.2.2 Amyloid tracers for amyloid plaque deposition

Amyloid PET imaging has contributed to a better understanding of Aβ processes in vivo. Amyloid PET tracers have been derived from several chemical classes, including thioflavin T derivatives (11C-Pittsburgh compound B (PIB), 18F-flutemetamol, 11C- AZD2184), stilbenes (18F- florbetapir, 18F-florbetaben and 11C-SB13), benzoxazoles (11C-BF-227) , aminonaphthalenes (18F-FDDNP) and benzofurans (18F-AZD4694) (Nordberg et al., 2010) (Figure 3, Table 1). 11C-PIB and 18F-FDDNP were the first amyloid tracers used to detect fibrillar Aβ (β-sheets) (Agdeppa et al., 2001; Klunk et al., 2004). 18F-florbetapir, 18F-florbetaben and 18F-flutemetamol have been approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) for clinical use to exclude non-AD.

Figure 3 Chemical structures of amyloid PET tracers.

Recent head-to-head comparative studies have suggested that the extent of cortical retention of 11C-PIB is similar to that of 18F-florbetapir (Landau et al., 2014), 18F- florbetaben(Villemagne et al., 2012a), 18F-flutemetamol (Vandenberghe et al., 2010) and 18F-NAV4694 (Rowe et al., 2013). The approved tracers 18F-florbetapir, 18F-

(19)

florbetaben and 18F-flutemetamol all have sensitivity and specificity between 80 and 90

% according to the results of phase III clinical trials (Curtis et al., 2015). Elevated amyloid loads have been detected in the frontal, parietal and cingulated cortices of patients with AD (Klunk et al., 2004; Kemppainen et al., 2007; Jack et al., 2009; Rowe et al., 2010), or MCI that converted to AD using 11C-PIB (Forsberg et al., 2008;

Nordberg et al., 2013), 18F-florbetaben (Rowe et al., 2008; Barthel et al., 2011), 18F- florbetapir (Clark et al., 2011). However, 10 - 30 % of cognitively normal adults also develop Aβ plaque deposition in the brain, increasing with age, according to

Alzheimer's Disease Neuroimaging Initiative (ADNI) and Australian Imaging,

Biomarker & Lifestyle Flagship Study of Ageing (AIBL) (Mintun et al., 2006; Rowe et al., 2010; Villemagne et al., 2013; Jack et al., 2014b).

There are strong correlations between in vivo PET results using 11C-PIB (Ikonomovic et al., 2008; Kadir et al., 2011), 18F-florbetapir (Clark et al., 2012), and 18F-

flutemetamol (Wolk et al., 2011; Rinne et al., 2012) and Aβ levels at autopsy. Amyloid tracers are able to detect senile plaques, diffuse plaques and CAA in the brain

(Ikonomovic et al., 2012). These tracers bind to fibrillar Aβ in AD brain with high affinity and specificity (Ni et al., 2013b). Recent biophysical modeling studies have provided insights on the tracer binding sites on Aβ fibrils. Molecular dynamic studies have identified six theoretical binding sites with Thioflavin T, PIB, BTA-1 (Groenning, 2010; Wu et al., 2011; Murugan et al., 2012) (Figure 4).

Figure 4 Binding sites of Thioflavin T on β-sheet structure of Aβ fibrils from molecular dynamic stimulation; six potential binding sites were identified, with a high-affinity binding site to surface β-sheet groove (1) (Murugan et al., 2012).

(20)

Dominantly Inherited Alzheimer Network (DIAN) and Alzheimer’s Prevention Initiative (API) have performed amyloid imaging studies in ADAD, and found

prominent tracer retention in both cortical and particularly subcortical brain regions in ADAD patients (Klunk et al., 2007; Koivunen et al., 2008; Villemagne et al., 2009;

Knight et al., 2010; Bateman et al., 2012; Benzinger et al., 2013; Fleisher et al., 2015).

Two exceptions, APParc (Schöll et al., 2012) and APP E∆693 (Tomiyama et al., 2008) mutation carriers showed low cortical 11C-PIB retention compared to sporadic AD patients, which has been attributed to ring-shaped plaques or oligomeric Aβ accumulation in the brain. Recent studies from API and DIAN suggest that the biomarker abnormalities start 15 years before the onset of symptoms, resembling the early changes in sporadic AD (Bateman et al., 2012; Fleisher et al., 2015).

Table 1 PET imaging tracers for detecting targets in AD brains

Target Ligands Reference

Aβ plaque

11C-PIB (Klunk et al., 2004)

18F-FDDNP (Agdeppa et al., 2001)

18F-Florbetapir (Clark et al., 2011)

18F-Florbetaben (Rowe et al., 2008)

11C-AZD2184 (Nyberg et al., 2009)

18F-AZD4694 (Cselenyi et al., 2012)

11C-BF-227 (Furukawa et al., 2009)

18F-Flutemetamol (Nelissen et al., 2009)

Tau

18F-THK-5105, 5117, 523 (Okamura et al., 2014)

11C-PBB3 (Maruyama et al., 2013)

11C-T808, 11C-T807 (Chien et al., 2014)

Glucose metabolism 18F-FDG (Herholz et al., 2002)

Microglia activation (TSPO)

11C-PK11195 (Cagnin et al., 2001).

11C-PBR28 (Okello et al., 2009).

18F-Fempa (Varrone et al., 2014)

Astrocytosis (MAO-B) 11C-DED (Carter et al., 2012)

TSPO = translocator protein; MAO-B= monoamine oxidase B.

1.2.3 Neuroinflammation

Currently used PET tracers for neuroinflammation mainly target astrocytosis and microgliosis in the brain (Jacobs and Tavitian, 2012). 11C-deuterium-L-deprenyl (11C- DED), a monoamine oxidase B (MAO-B) tracer, is an indicator for astrocytosis in the brain (Jossan et al., 1991; Fowler et al., 2005). Patients with MCI and high brain amyloid loads (high 11C-PIB retention) appear to have higher 11C-DED retention than

(21)

healthy controls or patients with AD (Carter et al., 2012; Choo et al., 2014). This suggests that astrocytosis is an early phenomenon during disease development. An autoradiography study using 3H-L-deprenyl and 3H-PIB showed different regional and laminar distribution patterns for reactive astrocytes and fibrillar Aβ in AD post-mortem brain slides (Marutle et al., 2013).

PET tracers targeting microglia activation, such as ligands for translocator protein (TSPO) 11C-PK11195, 11C-DAA1106, 11C-PBR28, and 11C-GE180 (Table 1) have demonstrated involvement of neuroinflammation at an early stage (Cagnin et al., 2001; Okello et al., 2009; Rupprecht et al., 2010; Varley et al., 2014). Recent finding of a polymorphism in the TSPO gene offers an explanation for the mixed binder observed in PET scans with the second-generation TSPO tracers (Owen et al., 2011;

Kreisl et al., 2013). The development of molecular imaging tracers to visualize neuroinflammation has been challenging, in part due to the changing/versatile states of astrocytes and microglia during the disease. Several novel neuroinflammation PET targets, such as the endocannabinoid receptor (CB 1R, 2R), cyclooxygenases (COX 1, 2), and arachidonic acid are under development (Jacobs and Tavitian, 2012; Varley et al., 2014).

1.2.4 Tau

A series of tau PET tracers have been developed for imaging tau pathology in AD and other neurodegenerative diseases, such as 18F-THK5105, 18F-THK5117, and 18F-T807.

The challenges of tau imaging in its intracellular location include the presence of different isoforms, its coexistence with Aβ, post-modification of the molecule and its presence in a series of neurodegenerative diseases (Villemagne et al., 2015).

1.3 REVISED SUGGESTED DIAGNOSTIC CRITERIA

As the pathophysiological changes in AD start many years before symptom onset, imaging biomarkers offer an invaluable tool for early differential diagnosis, and for finding links between the molecular pathology and the clinical phenotype (Villemagne et al., 2013). Since 2007, two sets of diagnostic guidelines for AD have been published by the International Working Group (IWG) and the National Institute on Aging-

Alzheimer’s Association (NIA-AA) (Dubois et al., 2007; McKhann et al., 2011; 2014).

The IWG-2 criteria categorized biomarkers currently used in MCI and AD into two categories: diagnostic pathophysiological marker and progression topographical

(22)

markers (Dubois et al., 2014). The IWG-2 criteria require the presence of an appropriate clinical AD phenotype (episodic memory impairment) and a pathophysiological biomarker that reflects AD pathology in CSF and amyloid

imaging. The NIA-AA criteria include different neuropathological stages: preclinical, MCI due to AD, and AD dementia; and use biomarker imaging, biofluids, and genetic information in the diagnosis. Both criteria recognize AD as a clinicopathological entity with a preclinical stage, and use AD biomarkers for the diagnosis (IWG) or as support for the diagnosis (NIA-AA) along with cognitive assessment (Morris et al., 2014). The use of specific biomarkers in molecular pathology increases the sensitivity and

specificity of the diagnosis. In addition to the difficulties associated with making an early diagnosis of AD, differentiating between AD and other types of dementia has been a challenge, since vascular dementia, frontal temporal lobe dementia, and Lewy body dementia have many overlapping factors with AD.

Patients with mild cognitive impairment (MCI) comprise a heterogeneous group that is at risk of developing dementia. MCI can be classified as amnestic or non-amnestic, according to which domains in the brain are impaired (Petersen et al., 2009). Some patients with MCI progress to dementia, while others remain stable for several years or revert back to cognitively normal (Petersen et al., 2009). Identification or prediction of those MCI patients who are at high risk of progression to AD or other types of

dementia would be valuable for early treatment. Biomarker diagnosis has been incorporated into new criteria for MCI due to AD (Dubois et al., 2007; Albert et al., 2011; McKhann et al., 2011; 2014).

1.4 MICROPET

The transgenic mouse model of AD mimics the amyloid pathology of AD and provides a tool for studying the time course of different pathological events in relation to

amyloid deposition. Transgenic mice with different mutations and genetic backgrounds have been associated with different types and magnitudes of pathology during life, which hindered interpretation of the results (Ashe and Zahs, 2010). An ideal transgenic animal model for AD should exhibit Aβ, tangles, synaptic loss, inflammation in the brain and memory problems. APPswe is one of the most commonly used animal

models; its phenotype includes Aβ plaques, vascular amyloid, gliosis, cognitive deficits and 5 to 6 times higher expression of APP. Other models, such as PDAPP, APP23 and tau transgenic mouse models PS19 display a different pathological signature. The use

(23)

of microPET imaging in transgenic animal disease model has provided valuable information for understanding the pathophysiology of the disease and for testing potential therapeutic strategies. A few PET studies have been performed in AD transgenic mice, using tracers for amyloid plaques, tau, inflammation, TSPO, glucose metabolism and cerebral blood flow (Higuchi, 2009; Teipel et al., 2011; Zimmer et al., 2014) (Table 2).

Table 2 PET imaging studies in transgenic AD mice and rats

Target Model Ligand Reference

Aβ plaques

APP/PS1 11C-PIB (Klunk et al., 2005;

Manook et al., 2012)

APPswe 11C-PIB (Toyama et al., 2005)

APPswe APP23 APPswe/PS1E∆9

11C-PIB (Snellman et al., 2013)

Rat model 18F-FDDNP (Teng et al., 2011)

APPswe 18F-AZD4694 (Jureus et al., 2010)

APPswe 18F-florbetaben (Rominger et al., 2013;

Brendel et al., 2014) APPswe 18F-flutemetamol (Snellman et al., 2012)

APP/PS1 11C-IBT (Yousefi et al., 2011)

APP23 11C-BF-145 (Okamura et al., 2004)

Tau PS19 11C-PBB3 (Maruyama et al., 2013)

PS19 18F-THK- (Okamura et al., 2013)

Microglia activation (TSPO)

APP/PS1 11C-PK11195 (Venneti et al., 2009;

Rapic et al., 2013)

APP23 11C-PBR (Ji et al., 2008)

Glucose

APPswe 18F-FDG (Luo et al., 2012)

APP/PS1 18F-FDG (Dubois et al., 2010)

3XTg 18F-FDG (Nicholson et al., 2010)

Aβ plaque +

Astrocytosis (MAO-B)

APPswe 18F-AZD4694 + 11C-DED Paper III

Aβ plaque + Microglia

activation (TSPO) PS19 APP23

11C-PIB + 18F-DAA1106 (Maeda et al., 2007;

2011)

Aβ plaque + Glucose APPswe 18F-FDDNP + 18F-FDG (Kuntner et al., 2009) APP/PS1 18F-AV-45 + 18F-FDG (Poisnel et al., 2012) Aβ plaque +

Cerebral blood flow

APP/PS1 APP23 15O-H2O + 11C-PIB (Maier et al., 2014)

TSPO = translocator protein; MAO-B= monoamine oxidase B.

(24)

1.5 TIME COURSES OF PATHOLOGICAL EVENTS IN AD In sporadic AD

The deposition of amyloid in the brain is estimated to start as much as 10-15 years before the onset of clinical symptoms. Changes in PET estimates of brain fibrillar Aβ deposition and CSF measurements of Aβ can be detected early in the prodromal stages of AD (Nordberg et al., 2010; Schöll et al., 2011; Jack et al., 2013; Villemagne et al., 2013; Donohue et al., 2014). PET results have indicated an increase in amyloid in patients with MCI, reaching a plateau on the development of AD (Engler et al., 2006;

Kadir et al., 2012). The temporal relationships between changes in Aβ, tau,

inflammation and other processes remain unclear. Amyloid deposition may precede changes in the neuronal injury marker CSF tau and FDG-PET-associated reduced glucose metabolism several years earlier than the manifestation of clinical symptoms in AD (Nordberg, 2011; La Joie et al., 2012) (Figure 3). Aβ accumulation is also

independent of hippocampal neurodegeneration (Jack et al., 2014a). The neuronal injury biomarker tau can provide results that are independent of Aβ, such as in patients with MCI with suspected non-amyloid pathology (SNAP) (Caroli et al., 2015),

suggesting an alternative pathological process in preclinical AD (Nordberg, 2011;

Chetelat, 2013).

The time course of inflammation in the development of AD remains unclear.

Microgliosis might be an early phenomenon during the disease. Early and declining astrocytosis, increasing amyloid-plaque deposition and decreasing glucose metabolism characterized the evolution of AD (Carter et al., 2012; Nordberg, 2014). Future

development of biomarkers and imaging tracers for tau, microgliosis and smaller forms of amyloid will shed light on the temporal changes of pathological events in AD.

In ADAD

Patients with early-onset AD caused by an ADAD mutation show clinical symptoms similar to those with sporadic AD, but with a younger age of onset and a positive family history. The recent finding of a protective APP gene indicate the role of APP processing in the disease etiology (Jonsson et al., 2012). A functional MRI study has indicated that the connectivity changes follow similar trajectories in sporadic AD and ADAD (Bateman et al., 2012; Fleisher et al., 2015). ADAD has a predictable age of onset and provides a model for understanding the sequence and pathological processes of sporadic AD. Longitudinal studies indicate that CSF changes appear 25 years before

(25)

the expected onset, followed by changes in amyloid imaging 15 years before, and changes in CSF tau and cerebral hypometabolism 10 years before (Fagan et al., 2014).

A longitudinal multi-tracer study using 11C-DED (reflecting astrocytosis) and 11C-PIB in ADAD patients showed an early increase in 11C-DED uptake, prior to the amyloid plaque deposition (Nordberg, 2014; Rodriguez-Vieitez E, 2015).

In healthy controls

A post-mortem study has indicated a certain frequency of Aβ lesions (Braak and Braak, 1997), and the presence of also age-related tau in the brains of people with normal cognitive function (Crary et al., 2014). Recent imaging and CSF biomarker studies have also shown that there is an age-related increase in cerebral Aβ and

neurodegeneration in people with normal cognitive function (Morris et al., 2010;

Villemagne et al., 2013; Jack et al., 2014b; Lim et al., 2014). It has been reported that amyloid abnormalities measured by PET and CSF amyloid levels may appear earlier than increases in CSF tau (Mormino et al., 2014). It is suggested that primary age- related tauopathy is common in human aging (Crary et al., 2014; Duyckaerts et al., 2015).The mechanisms for how people with neuropathological abnormalities maintain their cognitive function or cognitive reserve requires further study (Stern, 2012).

1.6 CURRENT AND FUTURE THERAPEUTICS

Current medications are symptomatic which include the AChE inhibitors (AChEIs) donepizil, galantamine, and rivastigmine, as well as the NMDAR antagonist

memantine. Drugs intended for a preventive or disease-modifying effects and which target different aspects of the pathological process are currently under development or in clinical trials. Current targets include slowing the production and increasing the clearance of Aβ, and eliminating existing Aβ pathology and toxicity (e.g. γ-secretase inhibitors, Aβ vaccines), as well as tau-targeted therapies such as tau phosphorylation, aggregation and tau antibody for clearance (Mangialasche et al., 2010; Corbett et al., 2012; Huang and Mucke, 2012).

Epidemiological data support the benefit of long-term nonsteroidal anti-inflammatory drug (NSAID) treatment: people receiving chronic treatment for an underlying inflammatory disease have a lower risk of developing AD (McGeer and McGeer, 2007). However, the results of large clinical trials of NSAIDs in patients with AD were not positive (Breitner et al., 2011), possibly because of the existence of particular

(26)

disease stage windows during which an anti-inflammatory strategy could be the most effective. Treatment with an α7 nAChR agonist has recently shown cognitive and neuroprotective benefits in a Phase IIb clinical trial for AD (Hilt et al., 2012).

One key question involves the optimal timing of specific treatment (at the preclinical or prodromal stages, or during mild AD). Recent phase III clinical trials with Aβ

antibodies, bapineuzumab (Salloway et al., 2014), solanezumab (Doody et al., 2014), and gantenerumab, however, have shown negative results (Karran and Hardy, 2014).

Theses failed late phase clinical trials indicated the importance of conducting trials with to the right patients, the right treatment/ dosage, the right time, and the right analytical method (Cook et al., 2014). Translation of results from preclinical to clinical situations and choosing a suitable patient population for the proposed mode of action are also prerequisites for developing therapeutic treatment. As the pathological events start years before the onset of clinical symptom, trials and early intervention are required in the preclinical stage. Biomarkers have already been applied to participant stratification, and as surrogate markers for efficacy in clinical trials. Trials such as the Dominant inherited Alzheimer Network trial (DIAN-TU) (Mills et al., 2013) in ADAD, the Anti- Amyloid Treatment in Asymptomatic Alzheimer’s (A4) prevention study in older individuals who may be at risk for AD (Sperling et al., 2014).

(27)

2 AIMS OF THE THESIS

In the present thesis, translational approaches including longitudinal clinical cohort, post-mortem sporadic and autosomal dominant AD brain tissue samples and a transgenic mouse model for AD have been used.

The specific aims were:

1. To evaluate the binding properties of various amyloid PET tracers; and compare the regional distribution of fibrillar amyloid in sporadic, autosomal dominant AD and control brain.

2. To investigate the time course of amyloid plaque deposition and astrocytosis in an experimental animal model by employing in vivo microPET and in vitro

autoradiography and immunochemical staining.

3. To develop a prediction model for conversion of MCI to AD using FDG-PET and CSF biomarkers.

4. To test the hypothesis that Aβ assemblies bind to α7 nAChRs and form complexes in AD brain.

(28)

3 METHODOLOGY

3.1 ETHICAL CONSIDERATIONS

Papers I, II and V: Autopsy brain tissues from sporadic AD patients and controls, three APPswe mutation carriers and one PS1 M146V mutation carrier was provided by the brain bank at Karolinska Institutet, Sweden and the Netherlands Brain Bank, Netherlands Institute for Brain Research, the Netherlands. Permission to use autopsied human brain material in experimental procedures was given by the Regional Human Ethics committee in Stockholm, and the Swedish Ministry of Health (Dnr 523/98, 1998-06-08, and Dnr 2011-91/3/31). Autopsy brain tissues from one PS1 M146V and one PS1 E∆9 mutation carrier were provided by Department of Pathology, University of Helsinki, Helsinki, Finland (Egentliga Finlands Sjukvårdsdistrikt Dnr 117/2005).

Paper III: The biochemistry and molecular imaging studies in APPswe transgenic mice were approved by Stockholm Södra Djurförsöksetiska nämnd (S53/10, 2010-05- 08; S63/10, 2012-03-30 Swedish National Board for Laboratory Animals) and the Animal Research Committee at Karolinska Institutet.

Paper IV: The 18F-FDG PET imaging studies and CSF studies in AD and MCI patients and control subjects were approved by the Regional Ethics Committee of Stockholm, Sweden and the Isotope Committee of Uppsala University, Uppsala, Sweden (Dnr 03-195, 2003-09-05, Dnr 2007/284-31.1). All patients and their caregivers provided written informed consent to participate in the study, which was conducted according to the declaration of Helsinki and subsequent revisions. Ethical approval was obtained from the Regional Human Ethics Committee of Stockholm and the Faculty of Medicine and Radiation, Hazard Ethics Committee of Uppsala University Hospital, Sweden.

3.2 PATIENTS AND MODEL SYSTEM 3.2.1 MCI patients (Paper IV)

Eighty-three MCI patients were recruited from the Department of Geriatric Medicine, Karolinska University Hospital Huddinge, Stockholm, Sweden who have been

referred from the primary care centres due to memory problem. The patients

underwent comprehensive clinical examinations, electroencephalograms, CT/ MRI, CSF and blood analysis including APOE genotype, and neuropsychological testing,

(29)

and were followed up every 6 months [mean 44 (1.6 - 162) months]. The diagnosis of MCI was based on clinical criteria as defined by Petersen et al (Petersen, 2004):

memory complaints, objective memory impairment (1.5 SD below age-matched controls),normal general cognitive function, intact daily living, and not fulfilling the DSM-IV criteria for dementia. Subjects were diagnosed as MCI-stable (not-

deteriorating) or as MCI-AD (converted to AD) based on the DSM-IV criteria for dementia and the NINCSD-ADRDA criteria.

3.2.2 Post-mortem human brain tissue (Papers I, II, V)

Brain tissues obtained at autopsy from patients clinically diagnosed and confirmed as having sporadic AD cases by pathological examination according to NINCDS-ADRDA criteria and from healthy controls was provided by the Netherlands Brain Bank,

Netherlands Institute for Brain Research, the Netherlands, and the Brain Bank at Karolinska Institutet, Stockholm, Sweden (Papers I, II and V). Brain tissue from ADAD patients carrying APPswe, PS1 M146V or PS1 E∆9 mutations was provided by the Brain Bank at Karolinska Institutet, Sweden and University of Helsinki, Helsinki, Finland (Papers II). Brain regions were dissected, and the number of NPs and NFTs quantified according to routine neuropathological protocols for the Brain Bank at Karolinska Institutet.

3.2.3 APPswe mice (Paper III)

Male and female transgenic mice expressing the APPswe mutation (APPswe, Tg2576) were obtained by in-house breeding at the Karolinska Institutet animal care facility.

Wild-type (wt) littermates served as controls. All mice were housed under the same conditions with access to food and water ad libitum and a 12h light/dark cycle. The mice were sacrificed by cervical dislocation within two weeks of the final PET scans, and their brains were quickly removed. The right brain hemispheres were stored at –80

°C and used in homogenate binding and autoradiography assays. The left brain hemispheres were post-fixed with 4% paraformaldehyde (pH 7.4), transferred to a sucrose cryoprotectant for 24 h at 4°C, and frozen at –80 °C for immunohistochemistry investigation.

(30)

3.3 PET AND DATA ANALYSIS (PAPER III, IV)

11C-AZD2184 and 11C-DED PET in APPswe and wt mice (Paper III)

Fifteen APPswe and nine wt mice (both male and female) underwent PET/MRI or PET/Computed tomography (CT) scans using a nanoScan® small-animal imaging system (Mediso Ltd, Budapest, Hungary) at the Karolinska Experimental Research and Imaging Centre, Stockholm, Sweden. 11C-AZD2184 and 11C-DED were synthesized at the Karolinska PET Radiochemistry Laboratory with a radiochemical purity > 98% and specific radioactivity of 181 ± 211 and 373 ± 443 (means ± SD) GBq/μmol,

respectively. The mice were anaesthetized with 1.5% (v/v) isoflurane. 11C-AZD2184 (10.6 ± 1.8 MBq) and 11C-DED (10.5 ± 1.6 MBq) were administered by venous tail injections. Dynamic PET scans were acquired over 63 minutes in 3D list mode.

Data analysis

The PET data were corrected for decay and dead time, reconstructed into 25 time frames. PET images were co-registered to a 3D digital T2-weighted MRI mouse brain template and analyzed by region of interest (ROI) using a 3D mouse brain atlas in PMOD 3.0 (PMOD Technologies Ltd, Zurich, Switzerland). An averaged 21-33 min frame was used for quantifying the extent of 11C-AZD2184 and 11C-DED radiotracer uptake, and the averaged activity in each target ROI was subsequently corrected for injected radioactivity and average weight of the mice and expressed in standard uptake value (SUV) units. The cortex, bilateral hippocampus and cerebellum were selected for further evaluation. 11C-DED binding was quantified using a simplified reference tissue model (SRTM) in PMOD using the cerebellum as a reference, and expressed as non- displaceable binding potential (BPND) for each ROI.

FDG-PET in humans (Paper IV)

All patients with MCI underwent 18F-FDG PET at the Uppsala PET Centre, Uppsala Academic Hospital, Uppsala, Sweden. The scans were performed under resting conditions, in a darkened room and with the fasted (for 4 h) patients’ heads resting in quick-setting foam. The scanners GEM 2048-15 and GEMS 4096- 15WB (General Electric Medical Systems, WI, USA) had a spatial resolution of 6 mm and covered 100 mm, with 6.5 mm slice spacing, producing 15 tomographic slices. The Siemens ECAT EXACT HR+ scanner (CTI PET systems Inc.) has a field of view of 155 mm, producing 63 contiguous 2.46 mm slices with 5.6 mm transaxial and 5.4 mm axial resolution. A correction factor was used for different scanners. Images were

(31)

reconstructed from the data and corrected for tissue attenuation of 511-keV gamma radiation photons by employing an external 68GE source. Accumulation of FDG in the brain was monitored for 60 min. Venous blood was collected from the back of the hand for measurement of plasma levels of radioactive FDG and glucose during the period. The Patlak graphical procedure was used to calculate the CMRglc value (μmol/min/100g) including a lumped factor of 0.418 to correct for differences in the utilization of FDG and glucose.

ROI analysis for FDG-PET

For the GE scanner, a set of 46 ROIs was defined using a Scanditronix program (IDA, Images Display and Analyses GE 1994). All ROIs were paired for the right and left hemispheres, except for the pons and the whole brain. Cortical ROIs (1×3 cm) were defined in the frontal, frontal association, parietal, parieto-temporal and anterior/posterior cingulate cortices. ROIs encompassing the caudate nucleus were defined at the level of the thalamus. Two circular ROIs (1.5 cm in diameter) were defined in two slices in the pons and then linked. ROI data were normalized to the pons metabolic rate for consumption.

For the ECAT EXACT HR+ scanner, FDG-PET was co-registered and the samples were resliced to their individual T1 reference images using SPM8. A simplified digital probabilistic atlas, consisting of 24 cortical and subcortical regions was applied for each individual’s native T1 space. These atlases were multiplied by the corresponding binary GM mask, which generated a GM-specific digital atlas for each participant. Raw co-registered and resliced FDG-PET data for each patient were sampled using the same individual digital atlases previously created. CMRglc values were created for FDG-PET and each atlas region by dividing by the respective mean pons metabolic rate. Five ROIs (frontal, temporal, parietal, posterior cingulate and caudate) and a composite ROI (mean of the five ROIs) were used.

3.4 CSF MEASUREMENTS (PAPER IV)

CSF was obtained by lumbar puncture from non-fasting subjects between 8 and 11 am at the Geriatric Memory Clinic, Karolinska University Hospital, Stockholm, Sweden. All samples were centrifuged for 10 minutes at 3000 g and 4 ºC immediately after collection. The supernatant was aliquoted and stored at -80 ºC until analysis.

Total tau (t-Tau), p-Tau and Aβ1-42 levels in the CSF were measured using a

(32)

sandwich enzyme-linked immunosorbent assay (ELISA) (INNO-BIA AlzBio3 assay, Innogenetics, Ghent, Belgium).

3.5 NEUROPSYCHOLOGICAL ASSESSMENT (PAPER IV)

The neuropsychological tests included assessments of specific domains such as verbal abilities (similarities and information), visuospatial abilities (block design and Rey- Osterrieth copy), episodic memory (auditory verbal learning and retention after 30 min; Rey-Osterrieth retention after 30 min) and attention and executive function (Trail making tests A and B). All cognitive raw scores were z-transformed using reference data from healthy adults at the Geriatric Clinic, Karolinska University Hospital Huddinge, Stockholm, Sweden.

3.6 IN VITRO RECEPTOR BINDING STUDY (PAPERS I, II, III AND V) Saturation assays (Papers I, II, III and V)

Saturation assays using amyloid tracers were performed by incubating brain tissue homogenates (100 µg tissue) with 0.01-250 nM 3H-AZD2184, 0.01-250 nM 3H- PIB,0.01-250 nM 3H-florbetaben and 0.01-250 nMmethyl-3H-BTA-1 in 1 x PBS buffer (pH 7.4) for 2 hr at room temperature. Non-specific binding was determined in the presence of 1 µM BTA-1. Samples were run in triplicate and the specific binding was expressed in pmol/g tissue. The dissociation constant (Kd) and maximum number of binding sites (Bmax) were determined using non-linear regression models in

GraphPad Prism version 5.0 (GraphPad Software, Inc., La Jolla, CA, USA). The saturation data were fitted to one-site and two-site binding models. The F-test was used to compare and select the appropriate binding model. Scatchard plots were prepared using GraphPad Prism to display the saturation data.

Competitive binding assays (Papers I, II, III and V)

Competitive binding assays comparing different amyloid tracers were carried out by incubating human brain homogenates (100 µg tissue) with 1.0 nM 3H-PIB, 2.5 nM 3H- florbetaben, 1.5 nM 3H-AZD2184 or 2.0 nM methyl-3H-BTA-1 in the presence of unlabeled ligand (AZD2184, BTA-1, florbetaben, florbetapir, BF-227 or FDDNP) at concentrations ranging from 10-11 to 10-4 M (Papers II-V). Binding assays with 3H- AZD2184, 3H-PIB and 3H-L-deprenyl were carried out on cortical homogenates (50- 100 µg tissue), from APPswe mice (Paper III).

(33)

The inhibition constant (Ki) and the percentage of displacement were determined using non-linear regression one-site and two-site binding models derived from the Cheng-Prusoff equation in GraphPad Prism. The F-test was used to compare and select the appropriate binding models. The three-site binding of 3H-PIB/AZD2184 in the frontal cortex of ADAD brain was analyzed by additional fitting using a two-site binding model in GraphPad Prism with binding data for AZD2184 concentrations ranging from 10-9 to 10-4 M (Paper II).

In vitro autoradiography (Papers I-III)

In Papers I and II, coronal frontal cortex brain tissue from three sporadic AD patients was cryostat sectioned (10 μm) at -20 °C and slide-mounted. Brain sections were thawed and dried for 15-30 min at room temperature followed by pre-incubation for 10 min in 1 x PBS buffer (pH 7.4 containing 1% BSA). Sections were incubated with 1.5 nM 3H-AZD2184, 1.0 nM 3H-PIB, 2.5 nM 3H-florbetaben and 2.0 nM methyl-3H-BTA- 1 in 1 x PBS buffer (pH 7.4) for 45 min. Non-specific binding was determined in the presence of 1 µM BTA-1. Some brain sections were pre-incubated, and then incubated with PBS buffer containing 10-6 M and 10-5 M p-nitrophenol, β-naphthol and

resveratrol.

In Paper III, autoradiography was performed by incubation of triplicate sagittal sections (10 µm) from APPswe and wt mouse brains with 3 nM 3H-AZD2184, 1.5 nM 3H-PIB and 10 nM 3H-L-deprenyl. Adjacent sections were incubated with unlabelled 1 µM BTA-1 or 10 µM L-deprenyl to determine non-specific binding.

Sections were rinsed 3 x 5 min in cold PBS, and briefly dipped in cold water before being air-dried overnight, placed together with calibrated tritium standards in cassettes, and exposed to Fujifilm BAS-IP TR 2040 imaging plates for 7 days. Phosphor image plates were processed using a BAS2500 phosphor imager (Fuji, Tokyo, Japan) at a resolution of 50 µm/pixel. The autoradiograms were analyzed with Multigauge software V3.0 (Fuji, Tokyo, Japan). The optical density values were converted into binding values (pmol/g tissue) calculated based on the tritium standards (Larodan Fine Chemicals AB, Malmö, Sweden).

(34)

3.7 IMMUNOCHEMISTRY (PAPER III)

APPswe and wild-type mouse brain sections were immunostained with mouse monoclonal Aβ42 (1: 200; Signet, USA), and polyclonal rabbit GFAP (1:500;

DakoCytomation, Denmark) antibody. Brain sections treated with either non-immune serum or serum without primary antibody served as controls. Sections were imaged sequentially at x10 and x20 magnification under light microscopy (Leica, Germany) with an image capture analysis system (ProgRes Capture Pro 2.8.8 software, JenOptik AG, Germany). The results were quantified using Image J software (NIH, Bethesda, MD, USA), and expressed as the number of immunopositive cells/mm2.

3.8 ELISA (PAPER I)

In Paper I, total Aβ40 and Aβ42 levels were quantified in post-mortem brain tissue from the frontal cortex and hippocampus of five AD patients and five control using commercial ELISA kits (Invitrogen, Camarillo, CA, USA).

3.9 STATISTICS

The data from Paper I were analyzed using SPSS (IBM SPSS Statistics, Version 22.0).

A T-test was used to examine the differences in mean age, education, and cognitive test scores at baseline, and a chi-square test was used to compare the male/female ratio and the frequency of APOE ε4 allele in these two groups. Logistic regression analyses were conducted to examine the feasibility of using demography, cognitive tests, APOE genotype, cerebral regional glucose metabolism, CSF biomarkers, and various

combinations of these to predict conversion to AD. A difference of -2 log likelihood (- 2LL) was used to compare the predictive ability of the variables. In Paper IV, data were analyzed using SPSS (IBM SPSS Statistics, Version 22.0) and GraphPad Prism version 5.0, 6.0. The non-parametric Mann-Whitney U test or one-way analysis of variance (ANOVA) with post hoc Dunnett test was used for comparisons between groups. Correlation analyses were carried out using the non-parametric Spearman’s rank order test. All values are shown as means ± SD. Significant differences between groups were set at *p < 0.05, **p < 0.01, ***p < 0.001.

References

Related documents

This project focuses on the possible impact of (collaborative and non-collaborative) R&amp;D grants on technological and industrial diversification in regions, while controlling

Analysen visar också att FoU-bidrag med krav på samverkan i högre grad än när det inte är ett krav, ökar regioners benägenhet att diversifiera till nya branscher och

dose of amphetamine as compared to placebo would cause an opioid release in the human brain reward system, measurable as a re- duction of the binding potential of the m-opioid

Our objective was to define the imaging profile of healthy elderly controls as a function of their cognitive trajectories and explore whether amyloid burden and white matter

The research on protein aggregation and amyloid formation is motivated by the fact that amyloid formation in tissue is harmful and associated with several debilitating diseases

The characterized transgenic mouse model suggests that CYP2C19 might be involved in the metabolism of endogenous substrates involved in important developmental processes

• To study cerebral white matter abnormalities and diffuse and excessive high signal intensities in a cohort of extremely preterm infants born in Stockholm

Other BRICHOS domains like Bri2 (Poska, et al., 2016), and Bri3 BRICHOS (data not shown) also form oligomers and reduce Aβ42 fibril formation in vitro (Paper III), and it is