Submitted for partial fulfilment of Doctor of Philosophy (PhD)

ANALYSIS OF THE SUBCELLULAR LOCALIZATION OF

PROTEINS IMPLICATED IN ALZHEIMER’S DISEASE

Anne Hwee Ling Lim

Supervised by Assoc Prof. Michael Lardelli and Dr. Joan Kelly

Department of Genetics and Evolution

School of Biological Sciences

The University of Adelaide Australia

1

DECLARATION

I certify that this work contains no material which has been accepted for the award of any

other degree or diploma in my name, in any university or other tertiary institution and, to the

best of my knowledge and belief, contains no material previously published or written by

another person, except where due reference has been made in the text. In addition, I certify

that no part of this work will, in the future, be used in a submission in my name, for any other

degree or diploma in any university or other tertiary institution without the prior approval of

the University of Adelaide and where applicable, any partner institution responsible for the

joint award of this degree.

I give consent to this copy of my thesis when deposited in the University Library, being made

available for loan and photocopying, subject to the provisions of the Copyright Act 1968.

The author acknowledges that copyright of published works contained within this thesis

resides with the copyright holder(s) of those works.

I also give permission for the digital version of my thesis to be made available on the web, via

the University’s digital research repository, the Library Search and also through web search

engines, unless permission has been granted by the University to restrict access for a period of

time.

Signed

…………………………………………………… Date……………………………

2

Table of Contents

Acknowledgements ................................................................................................................... 3

List of Publications contributed to during Ph.D. candidature ............................................. 4

Abstract ...................................................................................................................................... 5

List of abbreviations ................................................................................................................. 7

CHAPTER 1 .............................................................................................................................. 8

1.1 Introduction ........................................................................................................................... 9

1.2 Summary of chapters 2-6 and links between them ............................................................. 49

CHAPTER 2 ............................................................................................................................ 52

Research paper 1 ....................................................................................................................... 53

CHAPTER 3 ............................................................................................................................ 55

Research paper 2 ....................................................................................................................... 58

CHAPTER 4 ............................................................................................................................ 85

Research paper 3 ....................................................................................................................... 87

CHAPTER 5 .......................................................................................................................... 106

Research paper 4 ..................................................................................................................... 108

CHAPTER 6 .......................................................................................................................... 122

Research paper 5 ..................................................................................................................... 125

CHAPTER 7 .......................................................................................................................... 156

Discussion ............................................................................................................................... 157

References .............................................................................................................................. 167

APPENDIX I ......................................................................................................................... 180

3

ACKNOWLEDGEMENTS

I would like to give my thanks and appreciation to:

Associate Prof. Michael Lardelli, for providing me with this wonderful opportunity. Your

patience, knowledge and wisdom have guided and motivated me throughout my Ph.D. You

have been a continuous inspiration for the scientific process and the virtues of a critical

thinking lifestyle.

Dr. Morgan Newman, for your continuous support throughout my honours year till now. Your

advice and feedback has helped me greatly in producing this thesis.

Past and present members of the Genetics discipline for their support and advice over the last

few years. To Dr. Simon Wells and Dr. Zeeshan Shaukat, for advice on writing this thesis. To

everyone in the lab, particularly Dr. Hani Moussavi Nik, Swamynathan Ganesan and Tanya

Jayne, for adding a touch of fun to the lab atmosphere.

To Mom and Dad, for everything you have done for me and for all the support,

encouragement and belief you have in me. I could not have made it this far without you. To

my other family members especially my brother Jason, for the inspiration and support you

have given me.

To Dasrath, for putting up with me during the ups and downs, keeping me sane, and for

constantly reminding me that I can do it. Your understanding, patience, and faith in me have

been a great comfort.

Thank you.

4

LIST OF PUBLICATIONS CONTRIBUTED TO DURING PH.D

CANDIDATURE

Analysis of nicastrin gene phylogeny and expression in zebrafish.

Anne Lim, Seyyed Hani Moussavi Nik, Esmaeil Ebrahimie, Michael Lardelli.

Development Genes and Evolution, 2015

Differential, dominant activation and inhibition of Notch signalling and APP cleavage by

truncations of PSEN1 in human disease.

Morgan Newman, Lachlan Wilson, Giuseppe Verdile, Anne Lim, Imran Khan, Seyyed Hani

Moussavi Nik, Sharon Pursglove, Gavin Chapman, Ralph Martins, Michael Lardelli.

Human Molecular Genetics, 2014

Preparation of dissociated zebrafish cells for analysis of Psen1 truncation subcellular

localization.

Anne Lim, Morgan Newman, Michael Lardelli.

Zebrafish, 2015 (Under review)

5

ABSTRACT

Alzheimer’s disease (AD) is a neurodegenerative disorder involving neuropathological

changes including the presence of amyloid plaques in the brain. Amyloid plaques are

comprised mainly of the Aβ peptide, formed by the cleavage of the AMYLOID BETA A4

PRECURSOR PROTEIN (APP) by the enzyme complex γ-secretase. The catalytic

component of the γ-secretase complex is PRESENILIN (PSEN1 or PSEN2). Mutations in the

human PSEN1 gene have been identified in AD. Some mutations have been found to cause

frame shifts leading to premature stop codons that produce transcripts encoding truncated

PSEN1 protein. The subcellular location of the PSENs has been controversial, with studies

detecting these proteins in multiple areas in the cell. However recent research has shown that

the PSENs are enriched in the mitochondria associated membranes (MAM).

The zebrafish is a small freshwater fish that is a useful animal model for the study of human

diseases. Zebrafish have genes that are orthologous to human genes that play a role in AD,

including those that encode the components of the γ-secretase complex. However, the

zebrafish orthologue of the NICASTRIN gene has not been studied in detail. In Chapter II,

we characterize in zebrafish the orthologue of the human NICASTRIN (NCSTN) gene that

encodes a protein component of the γ-secretase. We demonstrate the spatial and temporal

expression level of zebrafish ncstn in embryos and various adult tissues. This work provides

the basic knowledge required to further the understanding of the role of Ncstn in the γ-

secretase complex using the zebrafish animal model.

Several truncations of the human PSEN1 and zebrafish Psen1 proteins appear to have

differential effects on Notch signalling and Appa cleavage. The basis for these differences is

still unclear. Chapter III examines the subcellular localization of fusions of truncations of

zebrafish Psen1 protein with green fluorescent protein (GFP) using a simple protocol to

obtain single cells from zebrafish embryos for microscopy analysis. We show that the

different truncations analysed appear to have similar distributions, suggesting that the

differential effects on γ-secretase substrates are likely not due to different subcellular

6

localisations of these truncated forms of Psen1. This chapter contributes a new method of

viewing a single layer of zebrafish cells with confocal microscopy.

The mitochondria-associated membranes (MAM) are a subcompartment of the endoplasmic

reticulum (ER) which interacts with mitochondria. The MAM is highly enriched in PSEN1

and PSEN2. Mitochondrion-ER appositions are increased in Psen1-/-/Psen2-/- mouse

embryonic fibroblasts (MEFs) and also in familial AD (FAD) and sporadic AD (SAD) patient

cells. These results, which were performed using mammalian cells, have not been studied in

the zebrafish animal model. Therefore, Chapter IV of this thesis examines the effect of

inhibiting zebrafish psen1 and psen2 activity on mitochondrion-ER appositions in 24 hpf

embryos. The findings in this study differ from those done in mammalian cells. In Chapter V

we use Western immunoblot analysis to determine if the MAM contains several proteins of

interest. The results of this study show that SORL1 protein, which has been linked to SAD, is

present in the MAM. The work in Chapters IV and V will help to give deeper insight into the

role MAM plays in both SAD and FAD.

Autophagy is a regulated catabolic process in which the cell digests cytoplasmic constituents

by lysosomal degradation. The role of autophagy in AD continues to be investigated, as

dysfunction in any of the stages of autophagosome formation could lead to

neurodegeneration. In Chapter VI we examine the ability of the antihistamine drug

Latrepirdine to induce autophagy in zebrafish larvae. The findings of this chapter indicate

Latrepirdine is capable of inducing autophagy in 72 hpf larvae, making it a suitable model to

study the mechanisms of this drug.

7

LIST OF ABBREVIATIONS

AD Alzheimer’s disease

AI Acne Inversa

AICD APP intracellular domain

CHO Chinese hamster ovary

CTF C-terminal fragment

DAP DYIGS and peptidase region

DRM Detergent-resistant membrane

EOAD Early-onset Alzheimer’s disease

ER Endoplasmic reticulum

FAD Familial Alzheimer’s disease

FTD Frontotemporal dementia

LOAD Late-onset Alzheimer’s disease

LTP Long-term potentiation

MAM Mitochondria-associated membrane

MCI Mild cognitive impairment

MEF Mouse embryonic fibroblast

MO Morpholino

NTF N-terminal fragment

PE Phosphatidylethanolamine

PS Phosphatidylserine

SAD Sporadic Alzheimer’s disease

TGN Trans-Golgi network

ROS Reactive oxygen species

8

CHAPTER 1

LITERATURE REVIEW

9

1.1. Introduction

Dementia is a term used to describe diseases that cause the progressive deterioration of mental

abilities such that it is severe enough to interfere with daily functioning. Alzheimer’s disease

(AD) is the most common form of dementia with 60-80% of dementia cases classified as AD [1,

2]. It is a neurodegenerative disorder characterized by progressive memory loss, in addition to

impairment of speech and motor ability, depression, delusions, hallucinations, aggressive

behaviour, and ultimately death (reviewed in [3]).The neuropathological hallmarks of AD are the

presence of neurofibrillary tangles and amyloid plaques in the brain. Neurofibrillary tangles

consist of the hyperphosphorylated forms of a protein called tau [4]. Amyloid plaques are

composed mainly of the Aβ peptide, formed from the cleavage of the C99 fragment of the

AMYLOID BETA (A4) PRECURSOR PROTEIN (APP) by the enzyme complex known as γ-

secretase. Besides these hallmarks, other abnormal features observed in AD patients are altered

lipid, cholesterol or phospholipid metabolism [5], aberrant calcium homeostasis [6] and

mitochondrial dysfunction (reviewed in [7]). Risk factors of the disease can be of a genetic or

non-genetic nature.

AD can be divided into two forms as defined by age, early onset AD (EOAD, <65 years) and late

onset AD (LOAD, >65 years) (reviewed in [8]. The majority of AD cases are of the sporadic

LOAD form (90-95%) [9]. EOAD cases only constitute 1-6% of all AD cases, of which

approximately 60% are familial (FAD), inherited from one generation to the next [8]. However,

even though EOFAD only makes up a small number of cases, a large body of research has been

10

focused on studying the defective genes linked to EOFAD as it has advanced the understanding

of the molecular pathology of AD. Genes linked to EOAD are known to have mutations that

cause AD. However, LOAD genes are associated as risk factors for developing AD but may not

result in the disease.

The Amyloid Cascade Hypothesis

There are several hypotheses regarding the pathogenesis of AD, with the amyloid hypothesis

being the most popular. It states that the accumulation of Aβ either via overproduction or lack of

clearance, leads to its oligomerization and deposition in the brain and ultimately to neuronal

dysfunction, degeneration and death [10]. The focus of the amyloid cascade hypothesis has

shifted from plaques to soluble forms of Aβ as the toxic forms that harm neurons and synapses.

Most research has focused mainly on how Aβ causes neurodegeneration. Some studies suggest

that one of the causes of neuritic dystrophy is the ability of Aβ peptides to promote oxidative

stress (reviewed in [11]). Aβ overexpression in transgenic mice, C. elegans, and cell cultures

results in increased biomarkers of oxidative stress [12]. In the AD brain, regions that are Aβ-rich

appear to have increased levels of protein oxidation [13]. Compared to normal brain, the AD

brain also shows decreased amounts of α-tocopherol, an antioxidant [14, 15]. However, it has

also been suggested that the presence of Aβ in the brain is a protective response to oxidative

stress [16]. Aβ can prevent lipid peroxidation at concentrations similar to those measured in

biological fluids [16] and can prevent metal-induced cell death in neuronal cell culture [17].

11

Other research suggests Aβ can bind to particular molecules on the neuronal surface [18-20] that

in turn affect signalling pathways. Aβ potentially interferes with long term potentiation (LTP)

[21]. Aβ appears to disrupt the activation of several kinases that are required for LTP and further

investigation indicated that Aβ binds to the insulin receptor, inhibiting it and multiple

downstream kinases important for hippocampal LTP [21]. This could lead to loss of dendritic

spines and synapses.

The γ-secretase complex, its components and substrates

The γ-secretase complex generates Aβ peptides of varying sizes but mainly forms Aβ40 and Aβ42.

In normal cells, Aβ40 is the more common species found, accounting for roughly 90% of total Aβ

[22]. Aβ42 only accounts for 10% of total Aβ in the cell [23]. However, Aβ42 is the predominant

peptide found in amyloid plaques, probably because it is more aggregation prone than Aβ40.

The γ-secretase complex is an aspartyl protease that proteolytically cleaves many type I

transmembrane proteins. These are single-pass transmembrane proteins with an extracellular N-

terminus and cytosolic C-terminus. Among the most well-known of the γ-secretase substrates are

AMYLOID BETA (A4) PRECURSOR PROTEIN (APP) and NOTCH proteins.

APP is a transmembrane protein with a large extracellular domain and a short cytoplasmic region

(reviewed in [24]). There are two pathways in which cleavage of APP occurs (Figure 1). In the

12

non-amyloidogenic pathway, APP is first cleaved in the extracellular juxtamembrane region by

α-secretase, releasing secreted soluble α-APP fragments, with a C83 fragment remaining

embedded in the membrane. The C83 fragment is then cleaved by γ-secretase to form a non

amyloidogenic P3 fragment and the APP intracellular domain (AICD). AICD translocates to the

nucleus and activates gene transcription [25]. In the amyloidogenic pathway, APP is cleaved by

BETA APP CLEAVING ENZYME (BACE), releasing soluble, secreted β-APPs and an

embedded C99 fragment. C99 is then cleaved by γ-secretase to generate smaller Aβ peptides of

different sizes, such as Aβ40 and Aβ42, while releasing AICD intracellularly [25]. Aberration in

proteolytic processing of APP leads to the formation of Aβ plaques in the brain and is associated

with Alzheimer’s disease [10].

Figure 1. Two pathways of proteolytic processing of APP. Schematic diagram representing cleavage of

APP. a) In the non-amyloidogenic pathway, APP is cleaved by α-secretase, releasing α-APP. γ-secretase

cleavage of the C83 fragment forms the P3 fragment and the AICD. b) In the amyloidogenic pathway,

BACE cleaves APP to form β-APP. C99 is cleaved by γ-secretase generating Aβ peptides and AICD

(Diagram taken from De Strooper et al. (2010)).

13

NOTCH proteins are cell surface receptors involved in cell fate decisions during development

[26, 27]. γ-secretase cleaves NOTCH at its intracellular domain when receptor-ligand binding

occurs. This cleavage releases the NOTCH intracellular domain (NICD) that similarly to AICD,

translocates into the cell nucleus and is involved in transcription regulation [26, 28].

The γ-secretase complex is the proteolytic complex critical for processing of many substrates

including APP [29-31]. The complex consists of four main components (Figure 2), the

PRESENILINS (PSEN1 OR PSEN2), NICASTRIN (NCSTN), ANTERIOR PHARYNX

DEFECTIVE HOMOLOG 1 (APH1) and PRESENILIN ENHANCER γ-SECRETASE

SUBUNIT (PSENEN). These four components interact in a stoichiometric ratio of 1:1:1:1 [32].

A substantial amount of literature has been dedicated to studying each component of the complex

in order to elucidate their function.

Figure 2. The four members of the γ-secretase complex within the lipid bilayer. PRESENILINS are

proposed to be nine-pass transmembrane proteins that are synthesized as approximately 50 kDa

holoproteins. Endoproteolytic cleavage occurs at the loop domain between transmembrane domains 6 and

14

7 to form an N-terminal fragment (NTF) of approximately 28 kDa and a C-terminal fragment (CTF) of

approximately 18k Da. The remaining proteins, NCSTN, APH1 and PSENEN (formerly PEN2) make up

the γ-secretase complex (Diagram taken from Tomita and Takeshi (2013)).

PSEN

PRESENILINS are transmembrane proteins that are core components of the gamma-secretase (γ-

secretase) complex, a multi subunit enzyme complex involved in intramembrane proteolysis of a

large number of transmembrane proteins. In humans, there are two PRESENILINS proteins,

PRESENILIN-1 (PSEN1) and PRESENILIN-2 (PSEN2) which share close to 67% sequence

identity and are encoded by the respective genes PSEN1 on chromosome 14, and PSEN2 on

chromosome 1[33].

The PRESENILINS are synthesized as approximately 50kDa holoproteins with 9 transmembrane

domains that undergo endoproteolytic processing to form an N-terminal fragment (NTF) of

approximately 28kDa and a C-terminal fragment (CTF) of approximately 18kDa in a 1:1

stoichiometry [34, 35]. Recent evidence points to the endoproteolysis of PRESENILINS being an

autoproteolytic event [36] with cleavage of the holoprotein being necessary to form the active γ-

secretase [36, 37]. Full length PRESENILIN is short lived, with a rapid turnover rate giving it a

half-life of approximately 1.5 hrs whereas cleaved fragments have a half-life of up to 24 hrs [38].

The function of the PRESENILINS in γ-secretase is believed to be that of the enzyme catalytic

core. PRESENILIN has two highly conserved aspartates at position 257 and 385 within

15

transmembrane domains 6 and 7 of the protein. Mutation of either aspartate residue results in loss

of endoproteolysis of PRESENILIN and abolishes the production of Aβ [39, 40]. Wolfe et al.

found that substituting either aspartate with alanine in PSEN1 prevented endoproteolysis of

human PSEN1 in Chinese hamster ovary (CHO) cells. They also observed significant reduction

in Aβ peptides and an increase in the amount of APP CTF compared to controls. This decrease in

Aβ and increase in APP CTF correlated with the amount of mutant holoprotein produced. No

changes in production of β-APP or α-APP were seen, indicating that the increased APP CTF

levels were due to reduction of γ-secretase activity and not due to increased cleavage by β- or α-

secretases [39]. Similar results were observed in PSEN2 mutants lacking either aspartate [40].

The PRESENILINS are best known for their role in the γ-secretase complex. However, they

appear to also have γ-secretase independent functions. PSEN1 has been shown to be involved in

lysosomal-dependent proteolysis (discussed later under the subheading “Autophagy and

Alzheimer’s disease”) [41]. The PRESENILINs in holoprotein form are able to form passive

calcium leak channels which allow a trickle of calcium ions to escape from the ER and into the

cytoplasm [42, 43].

NCSTN

The second protein to be identified in the γ-secretase complex, NICASTRIN (NCSTN) was

discovered through co-immunoprecipitation studies with PSEN1 [44]. NCSTN is a large 130kDa

type I transmembrane protein that can have a highly glycosylated ectodomain [44, 45]. NCSTN’s

role in the γ-secretase complex is unclear, but it may be involved in assembly and maturation of

16

the γ-secretase complex, with the ectodomain being necessary for γ-secretase complex maturation

but not its activity [46]. The NCSTN ectodomain contains a conserved glutamate 333 residue

(E333) that may have a significant role in the maturation of NCSTN via the secretory pathway

[46]. Furthermore, studies have shown that NCSTN might not take part in γ-secretase activity

[47]. γ-secretase retained activity in two independent NCSTN deficient mouse embryonic (MEF)

cell lines [47]. The trimeric γ-secretase component is similar biochemically to wild type γ-

secretase but is unstable in detergent solution compared to γ-secretase containing NCSTN,

signifying NCSTN’s main role is to stabilize γ-secretase [47]. Trimeric γ-secretase with PSEN1

mutants (S438P or F411Y/S438P) are also fully capable of proteolysing substrates [48]. It is

possible that the S438P mutation increased the stability of the γ-secretase complex, and therefore

reduced its dependency on NCSTN [48].

Alternatively, it has been argued that NCSTN’s role in the γ-secretase complex is in substrate

recognition [49]. The ectodomain of NCSTN has been shown to interact with γ-secretase

substrates, with the carboxylate side chain of the E333 residue playing a crucial role in

recognition of the N-termini of substrates [49]. The NCSTN ectodomain is the location of the

DAP domain (DYIGS and peptidase region) where E333 is also located. Also located in the

domain are five consecutive residues that make the conserved DYIGS motif. Interestingly, a

region downstream of the DAP domain mediates substrate binding but not γ-secretase complex

formation [50]. Furthermore, work by Xie et al. who studied the crystal structure of Nicastrin

from the eukaryote Dictyostelium purpureum found that the extracellular/luminal domain has a

large lobe which most likely constitutes the substrate binding site of Nicastrin, and a small lobe

which associates with the large lobe [51]. Interestingly, the conserved glutamate residue cognate

17

to human E333 and a Y293 residue that corresponds to Y337 of the DYIGS motif in human

NICASTRIN are located within the putative substrate-binding pocket of the large lobe, further

signifying the importance of the E333 residue and the DYIGS motif in substrate binding [51].

Figure 3. The topology of NCSTN. The large ectodomain contains the DYIGS motif and glycosylation

sites (indicated by the ovals). SP indicates the signal peptide, C indicates conserved cysteine residues, TM

indicated the transmembrane domain. The shaded areas indicate highly conserved domains between

human, murine, D. melanogaster and C. elegans NCSTN orthologues (Diagram taken from Yu et

al. (2013)).

APH1

ANTERIOR PHARYNX DEFECTIVE HOMOLOG 1 (APH1) contains seven transmembrane

domains with its N-terminal in the lumen and the C-terminal in cytosol [52]. In humans, there are

two APH1 homologs, APH1a and APH1b. APH1a has two splice variants, APH1aL and APH1aS

[53]. In mice, there is an additional homolog known as APH1c [54]. Deletion of APH1a in mice

causes embryonic lethality but deletion of APH1b or APH1c does not, implying that the different

APH1 isoforms perform different functions, with APH1a being important for embryonic

18

development [54, 55]. The actual role APH1 plays in the γ-secretase complex is not fully

elucidated although it is likely a scaffolding protein to which PRESENILIN can bind to [56].

Figure 4. The structure of APH1. APH1 is a seven transmembrane protein with its N-terminal facing the

lumen and the C-terminal in the cytosol (Diagram modified from Kimberly et al. (2003)).

PSENEN

PRESENILIN ENHANCER γ-SECRETASE SUBUNIT (PSENEN), formerly known as PEN2,

is a 101amino acid, double membrane spanning protein [57]. It binds preferentially to the

PRESENILINS in the complex, and endoproteolytic cleavage of the PRESENILINS to form NTF

and CTF requires the presence of PSENEN [58]. PSENEN appears to be important for

stabilization of the PSEN fragments and in maturation of the complex [59].

19

Figure 5. The structure of PSENEN. PSENEN is a double membrane spanning protein with both

terminals believed to face the cytosol (Diagram modified from Kimberly et al. (2003)).

Formation of the γ-secretase complex

The endoproteolytically cleaved PSEN1 protein along with the three other components, NCSTN,

APH-1, and PSENEN form the active enzyme complex. The assembly of the γ-secretase complex

is believed to begin at the endoplasmic reticulum (ER) [56]. It is believed immature NCSTN and

APH-1 form the scaffolding to which PSEN1 holoprotein binds, forming a trimeric intermediate

[56]. Structural analysis points to the NCSTN-APH1 complex binding to the CTF of PSEN1 [60-

62]. PSENEN is added to the intermediate, after which NCSTN undergoes a conformational

switch, resulting in the release of γ-secretase from the ER [56]. A structural prediction approach

modelled on the crystal structure of the PSEN1 archaeal homologue PSH indicates PSENEN

binds transmembrane 4 of PSEN1, which lies opposite to where NCSTN and APH1 bind in the

complex [63]. Maturation of the complex occurs further down the secretory pathway, via

endoproteolysis of PSEN to its NTF and CTF [64].

20

Genetic risk factors associated with familial Alzheimer’s disease

Mutations in three genes have been identified to cause FAD. These are the APP [65], PSEN1 [66]

and PSEN2 genes [33]. Of the three proteins, the focus has been on PSEN1 due to the large

number of mutations in the PSEN1 gene that have been linked to dementias such as Alzheimer’s

disease and frontotemporal dementia [67], the familial form of the heart condition dilated

cardiomyopathy [68], and the chronic inflammatory skin condition known as Acne Inversa [69].

Mutations in APP

The earliest reported APP mutation was in 1991 [65]. Since then upwards of 30 different

missense APP mutations have been described, most of which are pathogenic. APP mutations

located at the β and γ-cleavage sites of APP increase the Aβ42 to Aβ40 ratio [70].

Mutations in PSEN1

There are currently more than 170 known mutations of PSEN1

(http://www.molgen.ua.ac.be/admutations). These mutations are believed to affect APP

processing and increase the Aβ42: 40 ratio. Of these, most are missense mutations that do not

cause truncation or affect the C-terminal sequence of the protein [71]. However, aberrant splicing

of PSEN1 has been linked to disease progression in some cases of AD. Some mutations affect

splice sites, forming mRNAs lacking one or more exons. The L271V mutation was found in a

21

familial pedigree with FAD and encodes for a substitution of valine for leucine at codon 271,

within exon 8 [72]. The mutation appears to increase exclusion of exon 8 from mRNAs obtained

from AD brain and gives increased Aβ42 production in COS-7 cells [72].

A deletion of exon 9 (Δ9) from PSEN1 transcripts from AD has also been observed [37]. The

loss of exon 9 in PSEN1 transcripts can be due to a mutation affecting the splice acceptor site of

exon 9 [73], or a genomic deletion of exon 9 [74]. Loss of exon 9 in PSEN1 causes a high level

of Aβ42 production in AD brain [37].

Mutations have also been found that cause a frame shift leading to a premature stop codon and

transcripts encoding truncated protein. The PSEN1 delta 4 mutation was found in 2 patients with

EOAD [34]. It causes a 1 base pair (bp) deletion within the splice donor site of intron 4, resulting

in transcripts which lack part, or all of exon 4 and causes a shift in reading frame leading to a

premature stop codon. Normal full length transcripts, and full-length transcripts with an

additional TAC triplet coding for insertion of an amino acid after amino acid 113 are also formed.

A PSEN1 mutation has also been identified in frontotemporal dementia (FTD), a form of

dementia characterized by changes in behaviour and personality, impaired thinking and language

difficulties (reviewed in [75]). Among the neuropathological symptoms observed in FTD is the

atrophy of the temporal and frontal lobes [75]. The mutation G183V encodes a change from G to

T at codon 183, at the splice junction of exon 6 and intron 6 [67]. RT-PCR of a cDNA clone

22

sequence revealed transcripts lacking exon 6 or exons 6 and 7, with a shift in the reading frame

encoding a premature stop codon downstream of exon 5 [67]. Full-length transcripts with the

altered sequence are also produced [67].

Recently a PSEN1 mutation was found in one family suffering from familial Acne Inversa (AI), a

chronic inflammatory disease that affects the hair follicles [76]. This mutation caused a single bp

deletion in PSEN1 leading to a frameshift mutation [76]. Transcript expression of the mutant

allele was reduced, most likely due to nonsense-mediated decay. Mutations were also found in

NCSTN and PSENEN. Surprisingly none of the AI affected individuals showed symptoms of AD

or dementia.

The effects of interference of PSEN1 splicing have been studied further by Nornes et al. using the

zebrafish animal model (see later for zebrafish animal model). Zebrafish have orthologues of

human PSEN1 and PSEN2, named presenilin 1 (psen1) and presenilin 2 (psen2) [71].

Interference with translation of the Psen1 protein or with splicing of the psen1 transcript was

performed using morpholino antisense oligonucleotides (MOs), which are patented

oligonucleotides with backbone modifications that make them more resistant to nuclease

digestion (reviewed in [77]). MOs bind to the translation start codon/5´- untranslated region (5´

UTR) or splice donor/acceptor sites of mRNAs via Watson-Crick basepairing to prevent the

translation or splicing machinery from binding the target mRNA [77].

23

Nornes et al. attempted to model human PSEN1 splicing mutations such as L271V and PSEN1Δ9

using MOs. Zebrafish embryos were injected with MOs targeting the splice sites of the psen1

mRNA in order to exclude the exons of interest [71]. However, it was found that MO blockage of

the splice acceptor site did not result in exon exclusion; rather it resulted in the inclusion of the

preceding intron and failure of correct splicing out of targeted exons from transcripts. This in turn

resulted in the truncation of the open reading frame (ORF). These aberrant transcripts were

capable of evading nonsense mediated decay and were translated into truncated forms of the

Psen1 protein [71].

Due to their high homology and to simplify discussion, the zebrafish psen1 exons have been

referred in terms of the human PSEN1 exon number. Interference with Psen1 translation results in

phenotypes that resemble a loss of Notch signalling, likely due to loss of function of the γ-

secretase complex [71]. Likewise, truncation after exons 6 and 7 also appeared to show loss of

Notch signalling. A reduction in Notch signalling is known to cause increased expression of the

gene neurogenin1 (neurog1) [71]. Increased expression of neurog1 was observed in zebrafish

embryos injected with MOs inhibiting translation or causing truncation after exons 6 or 7, further

confirming a loss of Notch signalling [71].

A phenotype seen in zebrafish embryos during interference in exon 7 or 8 splicing was expanded

brain ventricles [71]. This phenotype is not due to loss of the γ-secretase activity as use of the γ-

secretase inhibitor DAPT failed to give this phenotype [71]. Therefore the expanded brain

ventricle phenotype is likely due to an alternative molecular activity of Psen1. Loss of Psen1 also

24

interfered with Psen2 activity, as it was found that the number of dorsal longitudinal ascending

(DoLA) interneurons increased relative to controls. The increase of DoLA interneurons is a

phenotype seen only when Psen2 translation is blocked, while decreased Psen1 expression does

not have an effect [78]. RT-PCR to quantify the level of mRNA splicing found that normal levels

of Psen1 mRNA were present with only low levels of aberrantly spliced mRNA produced. This

indicates that truncation of the Psen1 protein after exon 6 or 7 acts in a dominant negative manner

to reduce Psen1 activity, with only a small amount of aberrant protein required to give the

phenotype [71]. To confirm the results observed with MOs, mRNAs encoding the truncated

proteins were synthesized and injected into embryos, and confirmed the MO observations. It is

hypothesized that these Presenilin truncations interfere with normal Presenilin function through

unknown mechanisms that require elucidation.

The subcellular localization of truncated forms of PSEN1 proteins is currently unknown.

Interestingly, a substitution of aspartate 257 to alanine in PSEN1 causes a change in PSEN1

distribution, favouring anterograde flow to post-Golgi compartments over retrograde flow from

the Golgi to the ER [79]. Likewise, truncations of PSEN1 could cause changes in its distribution

in the cell. Identifying their cellular localization may give insights into their dominant negative

effects.

The subcellular localization of NOTCH and APP in the cell differs slightly. For instance,

NOTCH is localized mainly at the plasma membrane, but can be found in endosomes [80, 81].

25

APP has been found mostly within the cell along the late secretory pathway, with a small amount

at the plasma membrane [24].

Risk factors associated with sporadic Alzheimer’s disease

Several environmental and dietary risk factors are known to associate with SAD. These include

old age, previous family history of dementia [82, 83], high cholesterol levels [84, 85], obesity

[86], and diabetes [87]. This section covers a more in depth look at some of these factors.

The association between cholesterol and plaque formation in the brain appears to be dependent

on the fraction of cholesterol present. A study on serum cholesterol has identified elevated

cerebral Aβ with higher fasting levels of low-density lipoprotein (LDL), the cholesterol linked to

atherosclerosis, and lower levels of high-density lipoprotein (HDL) [88]. Although animal model

studies have shown that a high cholesterol diet does increase Aβ plaque formation [89-91] the

mechanism linking cholesterol with Aβ is not fully understood.

It has been suggested that the link between obesity and AD is through metabolic abnormalities

such as type 2 Diabetes mellitus, a disorder of glucose metabolism [92]. Insulin can be

transferred across the blood-brain barrier, where it can bind to receptors in the brain and

modulates cognitive function (reviewed in [93]). Large numbers of these receptors are located in

areas of the brain involved in memory such as the hippocampus and cerebral cortex (reviewed in

26

[94]). The presence of insulin increases the expression of the insulin degrading enzyme (IDE),

which degrades insulin as well as Aβ. Therefore insulin deficiency results in the accumulation of

Aβ peptides in the brain [95].

Recent work has indicated a link between periodontal disease and AD. Periodontal disease is an

inflammatory condition of the tissues supporting teeth that is caused by polymicrobial bacteria

present in the mouth. The host immune response against the pathogens gives rise to systemic

inflammation [96]. In the study, post mortem AD brain tissue was found to contain periodontal-

related bacteria, but not control brain tissue [96]. The hypothetical mechanism suggests that

bacteria transferred into the brain could cause vascular inflammation which would exacerbate AD

features [96].

Metal toxicity also appears to associate with SAD. Cortical iron elevation has been a feature seen

in AD brain and could be a contributor to oxidative damage in the brain [97]. Similarly,

aluminium toxicity has also been linked to SAD. Although its mechanisms are unclear, there is

evidence that aluminium is able to increase the leakiness of epithelial and endothelial barriers,

which in turn may allow its transfer from the blood to the brain [98].

27

APOE

Several genome wide association studies have been undertaken to try and identify genetic risk

factors of late sporadic LOAD [99]. One genetic risk factor that has been linked to a majority of

LOAD cases is the inheritance of one or more ε4 alleles of the APOLIPOPROTEIN E (APOE)

gene encoding the APOE4 protein which is involved in lipid transfer [99, 100]. The brain is

abundant in lipids (reviewed in [101]). However, lipids are insoluble and must be coated with

apolipoproteins such as APOE before they can be transported via receptor mediated endocytosis

[102, 103] between lipid producing cells such as from glial cells to neurons [104]. APOE has

three isoforms, with the most common allele being APOEε3 followed by APOEε4 and APOEε2

[102, 105].

Although it is known that APOE plays a role in lipoprotein metabolism, the neuropathological

pathway by which it is involved in AD has remained unclear [105]. Currently, there are several

hypotheses concerning the role of APOE ε4 in AD. One such hypothesis suggests that APOE

facilitates the clearance of Aβ from the brain by interacting with receptors to transfer Aβ across

the blood brain barrier (BBB) [102]. Castellano et al. have shown that cognitively normal

individuals under the age of 70 with the APOE ε4/ ε4 genotype have lower levels of Aβ in the

cerebrospinal fluid (CSF) compared to those with the ε2/ε3 or ε3/ ε3 genotypes [102]. This

indicates that, in individuals with the ε4/ ε4 genotype, Aβ tends to accumulate in the brain and

does not move into the CSF [102]. Using a dye that binds amyloid plaques in the brain, it was

also found that individuals with the ε4/ ε4 genotype bind more dye in the hippocampus and

28

interstitial fluid, compared to the other genotypes, further indicating accumulation of Aβ in the

brain [102]. On the other hand, other research appears to link APOE to Aβ deposition [103].

Studies using aged APPV717F AD mice models expressing APOE have shown significant Aβ

fibrilar deposits compared to controls [103]. However, APPV717F AD mice lacking APOE have

very little or no neuritic plaque development, suggesting that APOE is required for Aβ deposition

and fibril formation [103].

SORL1

Another gene identified by genome wide association studies to be associated with sporadic

LOAD is the neuronal sorting receptor SORL1, with multiple SNPs identified that significantly

associated with AD [106-108]. SORL1 is a 250 kDa type I transmembrane protein that is a

member of the low-density lipoprotein receptor (LDLR) family of APOE receptors and also the

vacuolar protein sorting 10 (VPS10) domain receptor family. SORL1 expression is widespread in

the brain, with most prominent expression in neurons of the hippocampus, some nuclei of the

brain stem and in Purkinje cells [109]. Subcellularly, SORL1 appears to be in the Golgi

compartment, plasma membrane and endosomes [110, 111].

The pathogenic nature of SORL1 in AD is not entirely clear. The discovery that SORL1 can

interact with APP and acts as a sorting receptor, therefore affecting its processing and trafficking

in the cell, indicates its importance in APP metabolism [111]. Two models have been suggested

29

for the mechanism by which SORL1 regulates APP metabolism. In the first model, SORL1

regulates the release of APP from the TGN to the cell surface [112]. SORL1 mediated

accumulation of APP in the TGN would in turn prevent APP cleavage by secretases and reduce

production of Aβ. The second model suggests it assists retrograde transport away from

amyloidogenic processing in the endocytic pathway where BACE cleavage of APP primarily

occurs [113].

The levels SORL1 present affect the production of Aβ produced within cells [111]. Increasing

SORL1 in cell lines appears to reduce processing of APP to produce Aβ while SORL1-deficient

mice show a 30% increase in Aβ levels compared to controls [111]. Some studies found SORL1

expression is reduced in the brain tissue of individuals with AD [111, 114]. SORL1 levels were

also reduced in cases of mild cognitive impairment (MCI) and their levels were intermediate

between no cognitive impairment and AD cases [115]. The reason for SORL1 reduction in the

brain is unclear but its reduction appears to relate to the severity of impairment of individuals.

Hypoxia, PS2V and reactive oxygen species

Evidence suggests a vascular component in the development of SAD. Studies have found that

cerebral blood flow declines with age. However, it is more pronounced in the AD brain [116],

with decreased density and fragmentation of microvasculature and Aβ deposition occurring in

blood vessels in the AD brain [117]. These changes to vasculature lead to hypoperfusion and

hypoxia of the brain, and can contribute to an increase in production of reactive oxygen species

30

[116]. Hypoxia has been shown to have an effect on BACE. Zhang et al. have shown that mice

neuroblastoma N2a cells exposed to acute hypoxia have increased expression of BACE1 mRNA

and protein, and increased production of Aβ peptides [118]. Hypoxia also increases expression of

PSEN2 [119], and PSEN1 [120] in rat and mice cells.

Hypoxia also induces HMGA1a, a non-histone DNA binding protein, which can bind to specific

sites on the fifth exon of PSEN2 pre-mRNA resulting in exon 5 exclusion from the pre-mRNA to

form the truncated isoform PS2V [121, 122]. PS2V has been found in the cerebral cortex and

hippocampus of SAD patients [121].

Studies relating to hypoxia have also been performed using zebrafish. Exposure to hypoxia

produces oxidative stress in zebrafish [122]. Interestingly, exposure of zebrafish to hypoxia leads

to increased expression of zebrafish bace1, psen1, psen2, and the paralogues of human APP,

appa and appb, at the embryonic developmental stage equivalent to 48 hours, and in adult

zebrafish brain [123]. These results obtained in the zebrafish animal model appear to be

consistent with results obtained using mammalian animal models and imply that the response of

these genes to hypoxia is evolutionarily conserved [123].

Alternative hypotheses to the Amyloid Cascade Hypothesis

The amyloid cascade hypothesis has dominated the field of AD research for more than 20 years.

This hypothesis proposes that toxic forms of Aβ are the main cause of AD [10]. The increased

31

levels of Aβ observed in AD patients compared to controls had given support to the hypothesis

[124, 125]. Based on the assumption that clearance of Aβ would improve cognition, several drugs

were designed targeting amyloid in the brain. Several drugs did in fact successfully remove Aβ

from the brain, however none of the drugs under clinical trials improved cognitive function

(reviewed in [126]). The failure of the clinical trials has cast doubt over whether Aβ is the causal

in AD, or perhaps may be associated with the disease but not the specific cause of the disease.

Furthermore, cognitively normal elderly individuals have been shown to have large deposits of

amyloid in brain tissue [127, 128] and this seems to indicate that AD neuropathology does not

necessarily associate with cognitive decline. Individuals with trisomy 21 (Down syndrome) who

possess three copies of the APP gene have elevated levels of the Aβ peptide and diffuse non-

fibrillar plaques but the prevalence of dementia was 70% by age 70, not the presumed 100%

prevalence for individuals growing old with Down syndrome ( reviewed in [129]). Most adults

(~95-98%) with Down syndrome in their 30s to 40s do not suffer from dementia [129].

Therefore, the presence of plaques does not always correlate with cognitive impairment.

Other abnormalities observed in AD patients such as altered lipid, cholesterol or phospholipid

metabolism [5], aberrant calcium homeostasis [6] and mitochondrial dysfunction (reviewed in

[7]) lack an apparent direct link to the amyloid cascade hypothesis. These less well-known

features of AD have prompted numerous alternative hypotheses focusing on cholesterol [130,

131], calcium [132], mitochondria [133], glucose [134], inflammation [135], tau [136] and more

to explain how each feature contributes to the pathogenesis of AD.

32

Subcellular localization of PSEN1

To better understand how all the features of AD fit together to bring about the disease, research

has focused on the subcellular level including identifying the localization of AD-related proteins.

The intracellular localization of APP and the components of its cleavage are of interest in order to

identify the cellular compartments in which Aβ is produced. The general consensus is that APP is

localized to the late secretory pathway and cell surface [24]. Previous studies estimated that 10%

of APP is found at the plasma membrane where α-secretase activity produces the p3 fragment,

while the majority of APP resides at the Golgi/trans Golgi network (TGN) [24]. APP has also

been identified in endosomes and lysosomes [137]. Similar to APP localization, BACE has been

located in the TGN, endosomes and the cell surface where it can come into contact with APP

[138]. BACE cleavage in these compartments can be linked to the generation of Aβ peptides in

the TGN and endosomes [24, 139, 140]. The presence of Aβ and p3 peptides is also indicative of

γ-secretase activity in plasma membrane, TGN and endosomes, however the location of its

catalytic component PSEN1, appears to be somewhat controversial. PSEN1 has been detected

abundantly in the early secretory compartments such as the ER or Golgi [141, 142]. Other studies

detected lesser amounts of PSEN1 in many other cellular locations such as the lysosomes [137],

mitochondria [143], centrosomes [144], interphase kinetochores [144] and nuclear envelope [79,

141, 144].

A recent study by Area-Gomes et al. sheds more light on the subject of PSEN1’s cellular

localization. They suggest that PSEN1 is highly enriched in the mitochondrial associated

33

membrane (MAM), a detergent resistant membrane (DRM), while PSEN2 appears to localize

exclusively at the MAM [145]. The MAM is a sub-compartment of the ER that is in close contact

with the mitochondria, allowing for physical interaction between the ER and mitochondria to

occur (Figure 3) [145].

The MAM appears to play several important roles. For instance, the MAM appears to regulate the

Ca2+ levels in mitochondria, via the presence of membrane receptors such as inositol-1,4,5

triphosphate receptors (IP3R) and sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) (reviewed

in [146]). MAM also contains proteins involved in lipid metabolism [145, 147]. The production

of phosphatidylethanolamine (PE) can occur in the ER alone but is mostly synthesized

collaboratively by the MAM and mitochondria. Phosphatidylserine (PS) is synthesized in the

MAM with the presence of the enzyme PHOSPHATIDYLSERINE SYNTHASE2 (PTDSS2)

[148]. PS can then be transferred to the mitochondria to synthesize PE [149]. The synthesis of

cholesterol to cholesteryl esters also occurs at the MAM [150].

The MAM is also the site of oxidative protein folding via the formation of an electron transport

chain involving two proteins, ERO1-LIKE PROTEIN ALPHA (ERO1L) and PROTEIN

DISULFIDE ISOMERASE (PDI) [146]. ERO1L generates disulfide bonds, which are then

transferred to PDI via oxidation of PDI by ERO1L [151-153]. PDI in turn catalyses the formation

of disulfide bonds within folding proteins [152, 153].

Besides PSEN1, the other components of the γ-secretase complex were also detected, along with

the γ-secretase substrate APP in MAM [145]. A fluorescence based γ-secretase assay using a

34

fluorogenic peptide substrate detected γ-secretase activity in the MAM while Western blotting

revealed cleavage of APP to form the AICD fragment, further demonstrating active γ-secretase

was present in the MAM [145]. This new finding may point to disruption of the MAM as playing

an important role in AD pathogenesis. Interestingly, APOE protein has been detected in the

MAM [154], implying that disruption of the MAM could be the link to the sporadic form of AD.

As altered lipid, cholesterol, or phospholipid metabolism and calcium homeostasis are among

some of the features seen in AD, and MAM is known to play a role in Ca2+ regulation as well as

lipid and cholesterol metabolism, altered MAM function may be the key to understanding AD

pathogenesis.

The close interaction between MAM and the mitochondria may also be linked to mitochondrial

dysfunction seen in AD [7, 155]. Aβ is found in mitochondria [156], and it is possible that Aβ

produced by APP cleavage in MAM is transferred to mitochondria, contributing to mitochondrial

dysfunction. Mitochondrial damage leads to a decrease in ATP production while increasing

production of reactive oxygen species (ROS) (reviewed in [157]).

Area-Gomez et al. further explained why the PSENs were not identified in the MAM in previous

localization studies. The authors have suggested that the discrepancy in PSEN1 localization may

be due to the lipid raft nature of MAM that can result in its co-purification with bulk ER, Golgi,

endosomes, and mitochondria [155]. Due to its detergent resistant nature, the use of detergents

for permeabilization in immunocytochemical techniques would only visualize the epitopes in

MAM that are not embedded within the DRM [155]. This may explain why PSEN and the other

35

components of γ-secretase were not identified in this sub-compartment earlier [155]. The use of

proper markers also allowed for the unequivocal identification of the MAM [145, 155]. It would

be interesting to ascertain if other proteins that may play a role in AD are localized to the MAM

and determine how their localization contributes to AD pathogenesis.

More recent work indicates cholesteryl ester and phospholipid synthesis is elevated in MEFs

lacking PSENs and in both FAD and SAD patient fibroblasts [158]. The number of ER-

mitochondrial contacts was also increased in these cells, with increased lengths between ER and

mitochondria compared to controls [158].

γ-secretase inhibitor treatment with DAPT increased cholesteryl ester synthesis in various cells,

indicating that increased cholesteryl ester synthesis appears to be due to a decrease in γ-secretase

activity. However, the same treatment appears to have little effect on co-localization of

mitochondria and ER and only causes minor changes to phospholipid synthesis [158]. These

results imply that intraorganellar functions within MAM, and interorganellar interactions

between mitochondria and MAM, require functional PSENs either as a part of γ-secretase or

independently.

36

Figure 3. the MAM and mitochondria. The MAM is a lipid-raft like domain of the ER that is in close

contact with mitochondria, allowing for communication between both organelles to occur. The MAM is

enriched in anionic phospholipids, cholesterol, and sphingomyelin, compared to bulk ER (Diagram taken

from Schon and Area-Gomez (2012)).

Autophagy and its role in Alzheimer’s disease

Autophagy is a regulated catabolic process where the cell digests cytoplasmic constituents such

as long lived proteins, macromolecular aggregates and damaged organelles by lysosomal

degradation [159]. The metabolites in these lysosomes can be reused by the cell, and so the

autophagy pathway can be seen as a recycling pathway for the cell. Additionally, autophagy is

involved in the clearance of intracellular microbes from the cell [160, 161].

37

There are three major types of autophagy; macroautophagy, microautophagy and chaperone

mediated autophagy. In microautophagy, the formation of the autophagosome sequestration step

is not required. Instead invaginations of the lysosomal membrane directly engulf the substrate

(reviewed in [162]). Chaperone-mediated autophagy does not require the formation of vesicles.

Instead, specific chaperone proteins bind to substrate proteins that contain a KFERQ sequence

and direct them to the lysosome surface [163, 164]. Macroautophagy is the major subtype of

autophagy and involves the formation of a double membrane vesicle known as the

autophagosome that sequesters cytoplasmic constituents that require degradation. The

autophagosome fuses with a lysosome to form an autolysosome and lysosomal enzymes degrade

the contents of the autolysosome. The further focus of this chapter will be on macroautophagy,

hereafter referred to as autophagy.

The underlying molecular machinery that drives the induction, elongation and formation of

autophagosomes involves many different proteins. Under normal, nutrient rich conditions, the

polyprotein complex mammalian target of rapamycin (mTOR) complex 1 (mTORC1) interacts

with the ULK1 protein complex (ULK1, ATG13, ATG101, FIP200). Initiation of autophagy

occurs with mTORC1 inhibition or AMPK activation which results in mTORC1 dissociation

from ULK1 and leads to phosphorylation of specific residues of ULK1. Activation of ULK1

catalyses phosphorylation of other components of the ULK1 complex and leads to its assembly at

isolation membranes [165]. ULK1 phosphorylates AMBRA, a component of the BECLIN-

1/PI3K complex (VPS34, VPS15, ATG14, BECLIN-1) and leads to its relocation to the isolation

membrane from the cytoskeleton, resulting in the production of phosphatidylinositol 3-phosphate

38

(PI3P) [166]. PI3P binds PI3P effectors, the WD repeat domain phosphoinositide-interacting

(WIPI) family of proteins, to mediate the initial stages of autophagosome formation [167, 168].

Various other proteins that interact with BECLIN-1 are involved in inducing or inhibiting

autophagy. For example, ATG14 and UVRAG (UV radiation resistance-associated gene) interact

with the PI3K complex in a mutually exclusive manner and compete for binding to the BECLIN-

1/PI3K complex in order to promote autophagosome formation [169-171]. BAX-

INTERACTING FACTOR-1 (BIF-1) which functions as a positive regulator of the PI3K

complex interacts with BECLIN-1 via UVRAG and likely generates membrane curvature [172,

173]. RUBICON (Run domain protein as BECLIN1 interacting and cysteine-rich containing) is a

negative regulator of autophagy [169, 170].

Following nucleation, the regulation of isolation membrane elongation and expansion involves

two ubiquitination-like reactions. In the first, ATG7 and ATG10 are required for the conjugation

of ATG5 to the ubiquitin-like molecule ATG12. This conjugate then interacts with ATG16L1 to

form the multimeric complex ATG12-ATG5-ATG16L1 which functions as the E3 ligase of LC3

(reviewed in [174, 175]). In the second reaction, the ATG12-ATG5-ATG16L1 complex interacts

with ATG3, ATG4 and ATG7 and mediates the conjugation of phosphatidylethanolamine (PE) to

the MICROTUBULE-ASSOCIATED PROTEIN 1 LIGHT CHAIN 3 (LC3)-I to form LC3-II

[174, 175]. LC3 is cleaved at its C-terminus by ATG4 to generate cytosolic LC3-I which has a C-

terminal glycine residue. This glycine residue is then conjugated to PE in a reaction that requires

ATG7 and ATG3, forming LC3-II. This leads to the translocation of LC3 from the cytoplasm to

39

both sides of the isolation membrane, and the isolation membrane closes to form the

autophagosome [176-178]. Most of the ATG12-ATG5-ATG16L1 complex dissociates from fully

formed autophagosomes, while LC3-II remains on the autophagosomes [179]. Levels of LC3-II

are increased at this point.

After the formation of the autophagosomes, its outer membrane fuses with a lysosome to form an

autolysosome. RAB7 is required for autophagosome-lysosome fusion [180, 181]. At this stage

ATG4 delipidates LC3-II on the outer surface of the autophagosome and converts it back to

cytosolic LC3-I [182]. RAB7 and LC3 interact with FYVE AND COILED-COIL DOMAIN

CONTAINING 1 (FYCO1) to form a complex that mediates microtubule plus end-directed

vesicle transport [183]. After fusion of the autophagosome with a lysosome, the lysosomal

enzymes degrade the intra-autophagosomal contents including LC3-II present on the inner

surface. Therefore the conversion of LC3-II to LC3-I by ATG4 and the degradation of the

luminal LC3-II by lysosomal hydrolases causes the level of LC3-II to decrease (reviewed in

[178]).

40

Figure 4. Autophagy induction and autophagosome biogenesis. Autophagy is the cell’s mechanism for

turnover of cytoplasmic constituents including organelles. Many proteins are involved in the molecular

machinery that drives the induction, elongation and formation of autophagosomes. Initiation of autophagy

occurs with mTORC1 inhibition or AMPK activation leading to formation of the isolation membrane,

elongation of the isolation membrane and closure of the inner and outer membranes to form the double-

membrane autophagosome (Diagram taken from Nixon (2013)).

Autophagy and Alzheimer’s disease

The presence of autophagic vesicles (AVs) is scarce in neurons of the healthy brain [184, 185]

with several possibilities as to why this is the case. One possibility is that the baseline autophagy

in healthy neurons is naturally low. A second possibility is that autophagosomes that form are

quickly cleared and do not accumulate within the cell. To answer this question, Boland et al.

41

performed a study using cultured cortical neurons derived from rat pups in which they induced

autophagy using rapamycin [185]. The strong autophagy induction in neurons resulted in LC3

compartments that were found to also contain active cathepsin D, a lysosomal protease indicative

of autophagosome-lysosome fusion, forming autolysosomes. The rapamycin-treated cultures

contained greater numbers of autolysosomes than autophagosomes when compared to controls

[185]. These observations indicate healthy neurons can initiate and sustain autophagy induction

but that accumulation of autophagosomes does not occur as these are converted to autolysosomes

and rapidly cleared from the cell.

In the AD affected human brain, the presence of AVs was frequently observed compared to

controls lacking AD pathology. The AVs appear to contain undigested or partially digested

cytoplasmic constituents [184]. The AVs appear to be mostly autolysosomes and are the main

organelles contained in giant neuritic swellings. Therefore autophagosomes likely can fuse with

lysosomes but cannot eliminate substrates properly [184, 186]. Interestingly, when Boland et al.

interfered with the clearance of autophagosomes by inhibition of cathepsins using inhibitors in

the cell cultures, they observed an increased number of LC3 compartments considerably

exceeding the numbers seen in rapamycin-treated cultures. Many of the AVs appeared to be

autolysosomes containing undegraded materials, resembling the AD pathology in human brain.

Thus inhibiting lysosomal proteolysis consequently leads to an accumulation of AVs in neurons

without altering induction of autophagy [185] and may point to disruption in the lysosomal

degradation step as critical in AD pathology. A critical finding has linked the AD genetic risk

factor PSEN1 to the lysosomal turnover of autophagosomal and endosomal protein substrates

[41]. The PSEN1 holoprotein physically interacts with the v-ATPase V0a1 subunit in order for it

42

to be N-glycosylated in the ER before it can be delivered to the lysosome [41]. The function of v-

ATPase is as a proton pump that acidifies newly formed autolysosomes in order for activation of

cathepsins to occur. Of particular significance was the finding that AD-related PSEN1 mutations

affect the targeting of the v-ATPase to lysosomes in fibroblasts from AD patients and cause the

impairment of autophagy [41].

Altered lysosomal proteolysis may not be the only potential contributor of AD, as dysfunction in

any of the stages of autophagosome formation is likely to cause neurodegeneration. For example,

the elimination of basal neuronal autophagy by loss of ATG5 expression in neural cells is

sufficient to cause neurodegeneration in a mouse model along with behavioural abnormalities

that are often observed in mouse models of neurodegenerative diseases [187]. This loss of

autophagy in cells causes an accumulation of abnormal proteins followed later on by the

formation of inclusion bodies [187]. It has also been found that BECLIN-1 mRNA and protein

levels are reduced in the AD brain [188]. Likewise, transgenic mice expressing APP and a

heterozygous deletion of BECLIN-1 appear to have accelerated extracellular Aβ deposits,

enlarged lysosomes containing electron dense materials and neuronal loss [188], pointing to a

role for autophagy in APP metabolism within cells, with autophagy potentially playing a

protective role against AD.

Until recently, the origin of isolation membranes was unclear with the ER, mitochondria and

plasma membrane thought to be the organelles from which the membranes derive [189, 190].

Isolation membranes were found to emerge from an omega shaped subdomain of the ER called

43

the omegasome that is labelled by the protein DFCP1 [191]. One of the most interesting recent

discoveries is that the initial site of formation of the autophagosomes is at mitochondria-ER (i.e

MAM) contact sites in mammalian cells [192]. Under starvation conditions, ATG14L and ATG5

exclusively relocate to the mitochondria-ER contact sites [192]. Another protein that is involved

in omegasome formation, double FYVE DOMAIN-CONTAINING PROTEIN 1(DFCP1) also

relocalises under starvation conditions [192]. The knockdown of PHOSPHOFURIN ACIDIC

CLUSTER SORTING PROTEIN 2 (PACS2), a protein involved in mitochondria-ER contact

formation, or MITOFUSIN 2 (MFN2), which tethers the ER to mitochondria, leads to a decrease

of ATG14 and DFCP1 in MAM fractions [192]. Therefore any disruption to the MAM-

mitochondria contact site would inhibit relocation of autophagy related proteins important in the

formation of autophagosomes. Interestingly another protein that also relocated along with ATG14

and DFCP1 was SYNTAXIN17 (STX17) [192]. STX17 appears to direct ATG14 to the MAM,

and its knockdown in HeLa cells lead to arrested autophagosome maturation at the isolation

membrane formation step [192]. Overall, this points to the MAM as a very versatile subunit of

the ER that has a major role in autophagosomes formation, and disruption to the MAM could

potentially affect many cellular functions including autophagy.

There has been some focus on developing a treatment for the reduction of protein aggregates as a

viable strategy to delay or stop the progression of neurodegeneration, with promising results in

Drosophila and mouse models of neurodegenerative diseases [193-195]. In particular, long term

mTOR inhibition by the drug rapamycin lowered Aβ42 levels and prevented the cognitive deficits

that are associated with the AD-relevant PDAPP transgenic mouse model [195]. However due to

the side effects of rapamycin [196] it would be desirable to develop safer alternatives that can

44

increase the induction of autophagy or reduce accumulated abnormal proteins that cause neuronal

dysfunction and cell loss.

The zebrafish as a model organism

Transgenic rodent models of AD have been used extensively due to the similarity in their genome

and brain anatomy to humans, along with the availability of various behavioural tests developed

for rodents to reveal neural dysfunction (reviewed in [197]). However, generating transgenic

rodents can be costly and time intensive. This has led researchers to use various other animal

models as alternatives. One such model that has been gaining popularity is the zebrafish Danio

rerio.

The zebrafish is a small freshwater fish native to northern India. Both larval and adult fish are

used in neuroscience research (reviewed in [198]). The zebrafish is gaining popularity as a new

model organism due to a number of characteristics. Firstly, zebrafish produce large numbers of

eggs which develop rapidly ex utero allowing for easy viewing and staging of embryos along

with swift completion of experimental work. Secondly, zebrafish are translucent during early

larval stages, making them suitable for tracking the expression of fluorescently labelled proteins

or dyes for fluorescent microscopy. Thirdly, the embryos are easily handled due to their large size

and are hardy enough to withstand experimental manipulation such as microinjection of

45

oligonucleotides. With regards to brain morphology, the zebrafish brain is fairly comparable to

rodent models [198]. Besides this, adult zebrafish have a relatively long lifespan of up to 5 years.

The genetics of zebrafish

Zebrafish are teleosts (bony fish) whose evolutionary lineage separated from the tetrapod lineage

at around 450 million years ago [199, 200]. The common ancestor of the teleostei underwent an

additional round of whole-genome duplication known as the teleost-specific genome duplication

(TSD) [200]. However, completion of the sequencing of the zebrafish genome has revealed that

around 70% of human genes have at least one clear zebrafish orthologue [200] making zebrafish

good models for genetic studies.

Figure 5. Zebrafish embryonic development. Zebrafish embryos are transparent, making them suitable for use in fluorescence microscopy. The fast development of the larvae allows for faster completion of experiments. (Diagram taken from Dahm and Geisler (2006)).

46

There are several ways in which the zebrafish genome can be manipulated. The transient

expression of disease-causing proteins can be studied by injecting synthesized DNA or mRNA

directly into single cell embryos. Genetic manipulation of zebrafish embryos can be furthered by

the injection of morpholino antisense nucleotides directly into embryos to target RNAs. A

morpholino subunit is made up of a nucleic acid base, a morpholine ring and a non-ionic

phosphodiamidate backbone [77]. The difference between the altered backbone of morpholinos

and the phosphodiester backbone of DNA/RNA makes it resistant to digestion by nucleases for

up to 48 hours. Morpholinos do not cause the degradation of their target RNA sequence. Instead

they work via a steric blocking mechanism and bind to complementary RNA by Watson-Crick

basepairing [77]. This allows for the use of morpholinos in translation blocking by preventing the

translation initiation complex from binding at the translation start site. Therefore morpholinos

need to be designed to target the 5ˊ – untranslated region (UTR) or the start codon. Morpholinos

have also been used to block splicing by targeting the morpholino to span the exon-intron

junction of pre-mRNAs. The morpholino sterically blocks splicing machinery from binding,

allowing for the modification of normal splicing events [201]. Unfortunately, targeting a splice

donor or acceptor site can also lead to the full or partial inclusion of introns and subsequently

lead to premature truncation of open reading frames [71].

More recent tools for genetic modification involve the use of sequence specific endonucleases to

introduce mutations into targeted sites in the genome. The two systems currently of greatest

interest are the transcription activator-like effector nucleases (TALENs) and the clustered

regularly interspaced short palindromic repeats (CRISPR) – CRISPR-associated (Cas) (CRISPR-

Cas) 9 systems. Both systems work to induce double strand breaks (DSBs) at the genomic target

47

site, introducing insertions and deletions (indels) [202, 203]. These systems can be used for

several genetic applications such as the generation of loss of function alleles [202, 203], or DNA

integration into the zebrafish genome [204] .

Zebrafish models of neurodegeneration/Alzheimer’s disease

Zebrafish are currently being used as relevant models of human neurodegenerative disorders such

as Parkinson’s disease, Huntington’s disease and AD as zebrafish have orthologues of the

disease-causing genes of the respective diseases.

Zebrafish have genes that are orthologous to human genes that play a role in AD. There are two

zebrafish genes, appa and appb, that are co-orthologues to human APP [205]. Zebrafish

possesses orthologues of all of the mammalian γ-secretase components [206-208]. Transcripts of

all the zebrafish γ-secretase components are detected before zygotic gene activation, indicating

the mRNAs are contributed to early embryos maternally [206-208]. The BACE1 orthologue,

bace1 has also been found in zebrafish [123]. The MAPT gene which encodes for tau protein has

two zebrafish co-orthologues, mapta and maptb [209]. Zebrafish also possess co-orthologues of

the sporadic, late onset AD associated gene APOE, apoea and apoeb [210, 211].

Zebrafish psen1 and psen2 are expressed maternally, and show ubiquitous expression in early

embryos, indicating they are important for embryonic development [206, 207]. Zebrafish

48

embryos that are translationally blocked for Psen1 protein with the use of morpholinos are still

viable, but show aberrant somite formation and Notch signalling defects [71, 208, 212]. This is

similar to conditional knockout mouse models (reviewed in [213]), although Psen1- /- mice have

been shown to die during development [214]. When Psen2 is blocked in zebrafish embryos,

Notch signalling is noticeably affected, whereas in Psen2- /- mice there is only a minor phenotype

[215]. The loss of Psen2 in zebrafish embryos affects Dorsal Longitudinal Ascending (DoLA)

interneurons but loss of Psen1 does not, signifying that different cell types rely differently on the

activity of these two proteins [78].

Investigating autophagy in zebrafish

Zebrafish possess one homologue of mTOR (mtor) [216]. Transcripts of zebrafish homologues

Beclin-1 and Lc3 are also expressed at 0 hpf, although the conversion of zebrafish Lc3I to Lc3II

only occurs at 32 hpf onwards, in contrast to mouse embryos where LC3II is detected in oocytes

[217]. The delay of autophagy induction may be due to the fact that zebrafish homologues of

proteins related to autophagy in other organisms, such as Ulk1a, Ulk1b and Atg9 only show

transcript expression after 23 hpf [218].

Zebrafish are a good model in which to measure autophagic flux. Autophagic flux is the process

from synthesis of the autophagosome to the degradation of the cytoplasmic constituents in

autolysosomes, and provides a good indicator of autophagic activity. This can be done by

49

comparing the Lc3II to Lc3I ratio in the presence of lysosomal inhibitors [218]. Since zebrafish

can be easily manipulated pharmacologically by the addition of chemicals of interest directly to

tank water, they can be used to screen a variety of chemicals for inhibitors or inducers of

autophagy.

1.2. Summary of chapters 2-6 and links between them

AD is a complex disorder that is characterized mainly by two main neuropathological hallmarks;

amyloid plaques and neurofibrillary tangles. The familial and sporadic forms of AD appear to

share similar clinical features of the disease, their underlying processes are distinct. Much

research has focused on pathogenic mutations found in PSEN1, the most common mutation

causing FAD. Furthering the understanding of the PSENs role in AD could give greater

understanding to the mechanisms of both SAD and FAD. Additionally, the MAM sub-

compartment may also prove to be important in understanding the underlying pathology of AD.

The zebrafish is a versatile animal model for the study of human disease. Zebrafish possess genes

orthologous to those involved in Alzheimer’s disease in humans. In Chapter II of this thesis, we

characterize in zebrafish the orthologue of the human NICASTRIN (NCSTN) gene that encodes a

protein forming part of the γ-secretase complex involved in formation of the Amyloidβ peptide.

Zebrafish Ncstn is predicted to possess the conserved glutamate 333 residue and DYIGS motif of

human NCSTN that are important for its functions, indicating that zebrafish Ncstn likely retains

similar functions to human NCSTN. We determine the spatial and temporal expression level of

50

zebrafish ncstn in embryonic and adult tissues. This work helps to further the basic knowledge of

the zebrafish ncstn gene and is a published article in the journal Development, Genes and

Evolution.

Having established a better understanding of zebrafish Ncstn, the work went on to study another

γ-secretase component, Psen1. Various truncations of the human PSEN1 and zebrafish Psen1

proteins have been shown to have differential effects on Notch signalling and Appa cleavage

[219]. The basis for these differences remains unclear. To further understand the effects of

truncations of PSEN1, the work in Chapter III explores the subcellular localization of fusions of

truncations of zebrafish Psen1 protein with green fluorescent protein (GFP). A simple protocol is

developed to obtain single cells from zebrafish embryos for microscopy analysis. The results

indicate that the different Psen1 truncations analysed appear to have similar distributions. This

suggests that the differential effects of these truncated proteins on γ-secretase substrates are likely

not due to differences in their subcellular locations.

Following the work on Psen1 truncations, the next chapter focuses on a known subcellular

location in the cell where PRESENILINs are found. The ER contains a subcompartment known

as the mitochondrial-associated membranes (MAM), which interacts with mitochondria. The

MAM is highly enriched in the PRESENILIN proteins [145]. Mitochondrion-ER appositions are

increased in Psen1-/-/Psen2-/- MEFs and also in cells from FAD and SAD patients [158]. In

Chapter IV of this thesis we examine the effect of inhibiting zebrafish psen1 and psen2 activity

51

on mitochondrion-ER appositions in zebrafish embryos. Chapter V further examines the

potential presence of several proteins of interest in the MAM.

The MAM is the initiation site of autophagosome formation. The final work in this thesis now

shifts focus from the MAM to autophagy. There has been some focus on the development of

drugs which can safely and effectively reduce the accumulation of abnormal proteins within the

brain. One potential therapy would be to modulate the autophagy pathway to improve β-amyloid

clearance. In Chapter VI we examine the ability of the antihistamine drug Latrepirdine to induce

autophagy in zebrafish. Our results suggest Latrepirdine is able to induce autophagy in 72 hpf

larvae, similar to observations in yeast [220].

52

CHAPTER 2

Research Paper 1

Analysis of nicastrin gene phylogeny and expression in

zebrafish

53

STATEMENT OF AUTHOURSHIP

Analysis of nicastrin gene phylogeny and expression in zebrafish.

Development Genes and Evolution, 2015

Anne Hwee Ling Lim (Candidate/First author)

• Performed phylogenetic and synteny analysis

• Performed qPCR and WISH experiments

• Wrote manuscript.

Certification that the statement of contribution is accurate

Signed

Seyyed Hani Moussavi Nik (co-author)

• Supervised and contributed to qPCR and WISH experiments

Certification that the statement of contribution is accurate

Signed

Esmaeil Ebrahimie (co-author)

• Performed statistical analysis for qPCR results.

Certification that the statement of contribution is accurate

Signed

54

Michael Lardelli (co-author)

• Planned the research and supervised development of work.

• Editing of paper

• Microscopy imaging of zebrafish embryos

Certification that the statement of contribution is accurate

Signed

Lim, A., Moussavi Nik, S.H., Ebrahimie, E. & Lardelli, M. (2015). Analysis of nicastrin gene phylogeny and expression in zebrafish. Development Genes and Evolution, 225(3), 171-178.

NOTE:

This publication is included between pages 54 - 55 in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1007/s00427-015-0500-9

55

CHAPTER 3

Research Paper 2

Preparation of dissociated zebrafish cells for analysis of

Psen1 truncation subcellular localization.

56

STATEMENT OF AUTHOURSHIP

Preparation of dissociated zebrafish cells for analysis of Psen1 truncation subcellular

localization.

Zebrafish, 2015 (Under review)

Anne Hwee Ling Lim (Candidate/First author)

• Performed all experiments (except where noted) and wrote manuscript.

Certification that the statement of contribution is accurate

Signed

Morgan Newman (co-author)

• Performed preliminary injection of mRNA into embryos.

• Supervised development of work for Lim.

Certification that the statement of contribution is accurate

Signed

57

Michael Lardelli (co-author)

• Planned the research, supervised development of work and edited the manuscript.

Certification that the statement of contribution is accurate

Signed

58

Preparation of dissociated zebrafish cells for analysis of Psen1 truncation

subcellular localization.

Anne Lim1, Morgan Newman1 and Michael Lardelli1

1. Alzheimer’s Disease Genetics Laboratory, Department of Genetics and Evolution, School

of Biological Sciences, The University of Adelaide, South Australia, Australia.

Corresponding author: Anne Lim

Tel.: +61 8 8313 4863

Fax: +61 8 8313 7534

Email: anne.lim@adelaide.edu.au

59

Abstract

Truncated forms of PRESENILIN (PSEN) proteins are possibility involved in the

pathological mechanisms of a number of neurological diseases such as Alzheimer’s disease

and Pick disease of brain. Depending on their size, the truncated forms appear to have

differential effects on cleavage of various γ-secretase substrates but the basis of these

differences is not known. One possibility is that different sizes of truncated protein have

different patterns of subcellular localisation and so differentially affect γ-secretase substrates

localised in different cellular compartments. To test this we generated mRNAs encoding

fusions of enhanced GFP (EGFP) to the N-termini of various truncations of the zebrafish

Psen1 protein. We injected these mRNAs into fertilized eggs and then observed their

localisation using fluorescence microscopy. We developed a simple protocol, described here,

to obtain single cells from zebrafish embryos for microscopy analysis. Cells obtained by this

method successfully retain organelle-specific dyes and express fluorescently tagged proteins.

The EGFP-tagged truncated Presenilin1 proteins were observed to be uniformly distributed in

the cytoplasm but not imported into the nucleus. Cells appeared to lack EGFP fluorescence in

lysosomes although this may have been due to the pH sensitivity of EGFP and protein

degradation.

60

Introduction

The human PRESENILIN proteins, PSEN1 and PSEN2, are the catalytic components of the

γ-secretase complex, an enzyme complex that is required for the intra-membrane cleavage of

over 70 different protein substrates. Among its substrates are the NOTCH proteins and the

AMYLOID BETA A4 PRECURSOR PROTEIN (APP). The PSEN1 gene contains 10 coding

exons which encode an approximately 50kDa holoprotein with 9 transmembrane domains. It

undergoes endoproteolytic processing forming an N-terminal fragment (NTF) and a C-

terminal fragment (CTF) in a 1:1 stoichiometry [1, 2]. Cleavage of the holoprotein is

necessary to form active γ-secretase [3, 4]. PSEN1 has two highly conserved aspartates within

transmembrane domains 6 and 7 of the protein that are important for endoproteolysis and

catalytic activity.

Mutations in the human PSEN1 gene have been identified in Alzheimer’s disease (AD) and

frontotemporal dementia (FTD, reviewed in [5]). Some of these mutations have been found to

cause frame shifts leading to premature stop codons, producing transcripts encoding truncated

PSEN1 protein [6, 7]. A truncated form of PSEN2 known as PS2V has been found in the

cerebral cortex and hippocampus of AD patients suffering from sporadic AD (SAD) [8] and

its formation is hypoxia-dependent. Hypoxia induces HMGA1a, a non-histone DNA binding

protein, which binds specific sites on the fifth exon of PSEN2 pre-mRNA resulting in exon 5

exclusion from the pre-mRNA and forming the truncated isoform PS2V [8, 9]. In addition to

mutations causing AD, a mutation (G183V) affecting the splice donor site of exon 6 in

PSEN1 has been found in patients suffering from Pick disease of brain, a form of

frontotemporal dementia (FTD). Interestingly, this mutation does not appear to produce any

associated AD pathology. The mutant gene produces transcripts coding for the full length

protein with a missense mutation, as well as transcripts with frame-shifted open reading

61

frames that code for proteins truncated after exon 5 [6]. A mutation (P242LfsX11) in PSEN1

causing a frameshift in exon 7 resulting in a premature termination codon was found in

families suffering from familial Acne Inversa (AI), a disease affecting the hair follicles [10].

None of the individuals with familial AI suffered from familial AD (Wang, pers. comm.).

Interestingly, work done in the zebrafish animal model has shown that various truncations of

the zebrafish Psen1 protein can have differential effects on zebrafish Notch signalling and

cleavage of Appa (a co-orthologue of APP) [11]. The different truncations can suppress or

stimulate Appa cleavage but not Notch signalling and vice versa (Fig. 3) [11]. For instance,

injection of mRNAs coding for a truncation resembling human PS2V resulted in increases in

both Notch signalling and Appa cleavage, consistent with previous work indicating that PS2V

increases γ-secretase activity [12]. A truncation resembling that encoded by aberrantly spliced

transcripts from the G183V mutant allele had no significant effect on Notch signalling but

decreased Appa cleavage, consistent with the lack of AD pathology observed in patients [11].

Injection of mRNA coding for a truncation resembling that potentially produced by the

familial AI P242LfsX11 mutation resulted in increased Notch signalling but had no effect on

Appa cleavage [11]. This is also consistent with the finding that none of the individuals with

this mutation had familial AD and it has been suggested that familial AI may be due to

changes in NOTCH signalling that affect follicle biology [10]. The cellular/molecular basis

for the differential effects of different Psen1 truncations is unknown.

Fluorescent probes such as fluorescent proteins and dyes have become a common feature in

molecular imaging. The availability of fluorescent proteins has proven beneficial in biological

imaging using fluorescence microscopy, particularly when used as a marker of gene

expression or to visualize protein interactions and localization in intracellular structures of the

62

cell. Fluorescent proteins are commonly used as fluorescent tags by fusion to a protein of

interest [13]. A commonly used fluorescent protein is the green fluorescent protein (GFP) of

the jellyfish Aequorea victoria [14]. GFP fusion proteins have been successfully targeted to

specific structures within the cell (reviewed in [15]). Fluorescent dyes with organelle

specificity have also been useful in biological imaging. Many dyes can be used on live cells,

allowing for live cell imaging to identify the subcellular structure in which a particular protein

may reside [16-18].

Among animal models, the zebrafish is an excellent model for in vivo imaging using

fluorescent probes. The advantages of using the zebrafish include that they produce embryos

externally. The embryos are produced in large numbers and are optically transparent during

early development. Zebrafish embryos are also robust and amenable to genetic manipulation

such as microinjection of DNAs or mRNAs into the cell to produce transiently or stably

transfected embryos [19, 20]. However a drawback of imaging whole organisms including

zebrafish embryos is the difficulty in visualizing processes within single cells due to the

complex matrix of cells within the organism and technical limitations involving microscopy.

Imaging single cells can be done using cell lines but this requires continuous maintenance and

the cells may develop abnormal phenotypes with time due to their unusual environmental

conditions (reviewed in [21]).

Here we present a simple dissociation method for zebrafish embryos to obtain single cells for

viewing by fluorescence microscopy. We have tested organelle-specific dyes on cells

obtained using this method. We also have generated several fusions of GFP to truncated

forms of the zebrafish Presenilin1 (Psen1) protein using the polymerase chain reaction (PCR)

63

driven overlap extension method. We use the cell dissociation method to analyse the

subcellular localization of these truncated forms of Psen1.

64

Materials and Methods

Ethics statement and animal husbandry

All zebrafish work was performed under the auspices of The Animal Ethics Committee of

The University of Adelaide. Wild-type zebrafish were maintained in a recirculated water

system with a 14 hour light/10 hour dark cycle. Fertilized embryos were grown at 28˚C in

embryo medium (E3) [22]

Materials

Pronase E (Sigma Aldrich, MO, USA, Cat. No P6911), Laminin (Sigma Aldrich, Cat. No

L2020), Dyes: MitoTracker Red CMXRos (Life Technologies, Cat. No M-7512),

LysoTracker Red,DND-99 (Life Technologies, Cat. No L-7528), Hoechst 33342 (2'-[4-

ethoxyphenyl]-5-[4-methyl-1-piperazinyl]-2,5'-bi-1H-benzimidazole trihydrochloride

trihydrate) (Life Technologies, Cat. No H1399), mMessage mMachine T7 transcription kit

(Life Technologies, Cat. No AM1344), MinElute Gel Extraction kit (Qiagen, Limburg,

Netherlands, Cat. No 28604), Phusion High Fidelity DNA Polymerase (New England

Biolabs, MA, USA), Vectors: N1-EGFP plasmid (Clontech, Cat No PT3027-5), The

pcGlobin2 plasmid [20] was a kind gift from Professor Myungchull Rhee, Chungnam

National University, Korea. Plasmids containing the DNA encoding the H2B-RFP and

CAAX-GFP (referred to as membrane-GFP) were kind gifts from Professor Jonathan Clarke,

Kings College London, UK. Ringer’s solution: 116mM NaCl, 2.9mM KCl, 1.8mM CaCl2,

5.0mM HEPES pH7.2, E2 embryo medium: 15mM NaCl, 0.5mM KCl, 1mM MgSO4,

0.15mM KH2PO4, 0.05mM Na2HPO4, 1.0mM CaCl2, 0.7mM NaHCO3

65

Table 1. Primer sequences.

Primers were synthesised Sigma Aldrich, MO, USA.

Primer Primer sequence (5'-3') Notes EGFP forward CGGGATCCACCATGGTGAGCAAGGGCGAG The underlined bases are the Bam H1

restriction enzyme recognition site. The bases in bold are the Kozak sequence.

EGFP reverse TAAATCAGCCATCTTGTACAGCTCGTCCATGCC The underlined bases are complementary to the sequence of the 5' end of psen1 forward.

psen1 forward GAGCTGTACAAGATGGCTGATTTAGTGCAGAATGCT The underlined bases are complementary sequence of the 3' end of EGFP reverse.

psen1Δ4 reverse GGAATTCTTACTGCTGTCCGTCCTTCTG The underlined bases are the Eco R1 restriction enzyme recognition sites. psen1Δ5 reverse

GGAATTCTTACTTGTAGCATCTGTACTT

psen1Δ6 reverse GGAATTCTTACAAATAAATTAAGGAG psen1Δ7 reverse GGAATTCTTAGTAGACTGAAATAGCAGCGAGGA

psen1Δ8 reverse GGAATTCTTAGGAGTAGATGAGCGCTGGGA

psen1ΔFL reverse

GGAATTCTTATATGTAGAACTGATGGACG

Generation of EGFP - psen1 fusion cDNA constructs

cDNA encoding the zebrafish Psen1 protein fused in frame to the enhanced green fluorescent

protein (EGFP) was made using the modified polymerase chain reaction (PCR)-driven

overlap extension method which requires two PCR steps (Fig. 1) [23]. Briefly, primers were

designed such that a reverse primer for EGFP cDNA is complementary at its 5' end to the 5'

end of a forward primer binding to zebrafish psen1 cDNA. This allows fusion to occur

between the EGFP and Psen1 coding sequences during PCR amplification. The forward

primer for EGFP contained a restriction site for Bam HI while the reverse primer for psen1

contained an Eco RI restriction site at the 5' end. Table 1 lists the primers used. The EGFP

cDNA fragment was amplified using the forward and reverse primers for EGFP cDNA in the

initial PCR step. Similarly the psen1 cDNA fragment was amplified using forward and

reverse primers for psen1 cDNA. High fidelity Phusion DNA Polymerase was used and the

PCR parameters followed the manufacturer’s protocol for Phusion polymerase, with the

number of cycles kept to 15 to reduce the risk of nucleotide mis-incorporations. Each PCR

66

product was analysed on a 1% agarose gel and the size of each PCR fragment was confirmed

before excision from the gel using a clean razor blade and placement into a 1.5 mL tube. The

PCR fragments were purified using the QIAquick Gel Extraction kit following the

manufacturer’s protocol. The fusion of the two fragments was performed in the second PCR

step, where equal concentrations of the EGFP and psen1 fragments were added to the PCR

reaction mix along with the forward primer for EGFP and the reverse primer for psen1. PCR

parameters remained the same as during the initial PCR. The second PCR generated the

EGFP-psen1 fusion cDNA construct. The cDNA construct was gel-purified followed by

restriction with Bam HI and Eco RI, and ligation into pcGlobin2, a vector system which is

suitable for the transcription of in vitro capped mRNA as it contains three elements (the 5' and

3' zebrafish β-globin UTRs and a 30 mer poly (A) tail) that generate mRNAs with increased

stability and translation efficiency [20]. Next, the EGFP-psen1 fusion cDNA construct was

used as a template in PCR to obtain cDNAs encoding EGFP - psen1 truncations varying in

their number of exons. Each truncated psen1 cDNA construct was inserted into pcGlobin2.

mRNA encoding the various fusion proteins was made using the mMessage mMachine T7 in

vitro transcription kit following the instructions provided. The resultant mRNA was diluted in

RNase-free water.

Microinjection of zebrafish embryos

Fertilized embryos were collected from wild type zebrafish and placed in a petri dish

containing E3 medium. Embryos were injected with mRNA at the one cell stage before

cleavage occurred and then incubated at 28˚C. To determine if mRNA injections were

successful, the embryos were observed under a fluorescence dissection microscope for

chimeric protein expression at approximately 2.5 hpf.

67

Dye staining of zebrafish embryos

Once expression of a GFP-Psen1 fusion protein was confirmed, the embryos were placed in a

small petri dish and dechorionated using Pronase at a concentration of 0.5 mg/ml. We found

that the chorions were easily removed once embryos were in Pronase for approximately 4

minutes with gentle swirling of the petri dish. Embryos were removed when touching the

chorion with forceps resulted in an indentation. Care was taken not to leave the embryos in

Pronase for too long as this led to damage of the embryos. To remove the embryos from

Pronase, the petri dish containing the embryos was slowly submerged in a large beaker

containing water. The embryos were allowed to sink to the bottom of the beaker before 2/3 of

the water in the beaker was poured off and replaced with clean water. This was repeated three

times to ensure all Pronase was removed. The dechorionated embryos are very fragile and

cannot be exposed to air as this damages them. At 3 hpf, 12-15 embryos were incubated with

gentle rocking in Ringer’s solution with organelle-specific dyes. Table 2 lists the dyes and

conditions used.

Table 2. Organelle dyes.

Organelle stain Concentration Incubation period (min) Wash period (min) MitoTracker Red CMXRos 100nM 50 10 LysoTracker Red DND-99 75nM 30 -

Hoechst 33342 9.7μM (6μg/mL) Added prior to loading sample onto coverslip/coverglass -

Embryo dissociation and microscopy

At the end of the incubation period, the animal pole of each embryo was separated from the

yolk using a needle. The cell clumps of each animal pole were transferred to a 1.5 mL tube

and clean Ringer’s solution was added for the wash period if necessary. Hoechst 33342, a

68

blue fluorescent nuclear stain, was added to each tube to a final concentration of 6 μg/mL. To

dissociate the cell clumps, a 200μL pipette tip with the end trimmed was used to gently

pipette the cells up and down several times. Approximately 35-40 μL of the sample was

placed on the surface of a clean 22 X 60 mm coverslip using a trimmed 200 μL pipette tip. A

smaller 18 X 18 mm coverslip was then coated on all 4 edges with a small amount of silicone

grease before being gently placed over the sample to prevent the sample from drying.

Alternatively, glass bottom dishes can replace coverslips, with the sample directly added onto

the coverglass and imaged. Imaging was performed using a Leica TCS SP5 confocal

microscope system with sequential scanning for multiple fluorophores. Images were obtained

from a single optical plane.

69

Results

Testing of an embryo dissociation method for observing single cells

The embryo dissociation method we developed for obtaining single cells for microscopic

analysis is described in Materials and Methods. We first tested the dissociation method using

organelle specific dyes to observe the ability of the cells to retain the dyes following cell

dissociation. The dyes used were MitoTracker Red CMXRos, a rosamine derivative

containing a mildly thiol reactive chloromethyl moiety, which labels mitochondria, and

LysoTracker Red DND-99, an acidotrophic probe suitable for labelling lysosomes. We found

that the cells successfully retained the dyes after dissociation was performed on the embryos.

However we found that labelling of the cells varied, as some cells had low levels of

fluorescence intensity while other cells had noticeably higher levels (Fig. 2a). This mosaic

labelling of the cells likely occurs due to the differential penetration of the dye into embryos

with surface cells receiving most dye.

We next injected embryos with mRNAs coding for CAAX motif membrane-green fluorescent

protein (membrane-GFP) and histone 2B-red fluorescent protein (H2B-RFP). The CAAX

motif targets a protein to the plasma membrane [24]. The nuclear protein H2B is one of five

kinds of histone proteins which interact with chromatin to form the nucleosome [25, 26].

These constructs have been successfully used previously by Alexandre et al. following

injection into 64-128 cell stage embryos and imaging at 21-24 hpf [27]. We injected the

mRNAs at the one cell stage and observed the effect of the dissociation method on subcellular

localization of these proteins. The expression of H2B-RFP and membrane-GFP was mostly

confined to the nuclear region and cell membrane respectively, indicating that the cell

dissociation method does not noticeably affect the subcellular localization of these proteins

70

(Fig. 2b). However, mosaic expression of both proteins was observed, similar to that observed

for dye labelling of cell organelles. Cells showed varied levels of fluorescence intensity of the

RFP- and GFP-tagged proteins. This is likely due to uneven initial distribution of injected

mRNA. We also observed autofluorescence of any yolk still attached to cells when using the

488nm laser to excite GFP. Therefore the use of nucleus-specific dyes such as Hoechst 33342

or the injection of mRNAs encoding fluorophore-labelled nuclear proteins is necessary in

order to distinguish cells from yolk debris.

Cell survival was found to be approximately 3 to 3.5 hours in Ringer’s medium before cell

blebbing followed by apoptosis occurs. We did not test other types of media, but it may be

possible to increase survival time by using suitable growth media [28]. We found that not all

cells formed a single layer, as clumps of cells were still visible. However attempts to reduce

cell clumps by increased pipetting to increase dissociation caused high rates of cell

fragmentation as the cells are too fragile and could not withstand the increased force. Many

cells still maintained a spherical appearance indicating that they did not fully adhere to the

coverslip surface during the short timeframe. The use of coating substrates for coverslips such

as Laminin did not appear to increase cell attachment noticeably.

Subcellular localisation of different truncations of Psen1

The differential effects of the Psen1 truncations on Appa cleavage and Notch signalling could

be due to the different subcellular localizations of these truncated proteins, as well as

differences in substrate localizations within the cell. Thus far PSEN1 has been detected

abundantly in the early secretory compartments such as the ER or Golgi [29], and in lesser

71

amounts in other cellular locations such as the lysosomes [30], mitochondria [31],

centrosomes [32], interphase kinetochores [32] and nuclear envelope [29, 32, 33]. More

recent work has indicated that PSENs are highly enriched in the mitochondria-associated ER

membranes (MAM) [34]. We were interested to know the effects truncation of a protein

would have on its localization within the cell, specifically whether different truncations

localize to different organelles within the cell. Using the modified-PCR driven overlap

extension method, we created cDNA constructs of EGFP fused in frame to the full-length

zebrafish psen1 coding sequence as well as psen1 sequences possessing varying numbers of

exons, specifically psen1 truncated after exon 4,5,6,7 and 8 respectively (Δ> hereby denotes

“truncation after”). We chose to fuse the COOH-terminus of EGFP to the NH2-terminus of

Psen1 as the C-terminal end of EGFP is flexible and rarely requires the presence of a linker

for flexibility between the two proteins [35]. Additionally, Psen1 is a nine-pass

transmembrane protein, with its N-terminus facing the cytoplasm [1, 2]. Thus, placing EGFP

at the Psen1 C-terminus might affect its conformation within the membrane. Fusion of GFP at

the N-terminal of human PSEN1 does not appear to affect its activity [36]. The mRNA for the

various constructs was synthesised and injected into embryos at the one-cell stage.

We first injected mRNA encoding the EGFP–Psen1 full length fusion, but could not detect

noticeable fluorescence under the dissection microscope at 2.5 hpf. We have previously

observed that it is difficult to force ectopic expression of full-length Presenilin protein in

zebrafish embryos if endogenous protein expression is not blocked (Michael Lardelli,

unpublished observations). Next we injected mRNAs encoding the different EGFP-Psen1Δ>

constructs into embryos and stained them with MitoTracker dye before cellular dissociation.

A relatively low level of EGFP fluorescence intensity was detected throughout the cytoplasm

of the cell for the five different truncations and there appears to be overlap with the

72

MitoTracker dye (Fig. 4). The truncated proteins do not seem to localize to a specific

organellar region of the cell, although no importation into the nucleus was observed. In some

of the cells examined there appeared to be areas within the cytoplasm where a lack of green

fluorescence was detected (Fig. 4 and Fig.5, arrowheads), initially indicating that these areas

may not contain the chimeric protein. When we stained cells with LysoTracker dye, the areas

lacking fluorescence appeared to overlap with the LysoTracker dye stain (Fig. 5).

73

Discussion

Our dissociation method for zebrafish embryos is a simple and quick way to observe

genetically manipulated single cells under fluorescence microscopy. A benefit of using

embryos at the earlier time points used in this protocol is that cells can be mechanically

separated without the use of proteinases or calcium- and magnesium-free medium, which are

normally used when dealing with later stage embryos. The use of such harsh conditions may

affect cellular function and, consequently, protein localization within the cell. One

disadvantage of this protocol is that the cells do not survive for a long period of time.

However the use of suitable growth media such as Leibovitz medium which has been used

previously with zebrafish spinal neuron cultures [28] may increase survival time. Determining

the most suitable media will be a future task to improve the protocol. Increasing cell survival

time will allow more time for cells to fully adhere to the slide surface and should improve

imaging. The protocol described in this paper can potentially be used for a number of

applications such as determining the sub-cellular localization of proteins of interest,

determining the successful expression of a fusion construct or testing new organellar dyes.

We have successfully used this method to observe the localization of the EGFP-Psen1Δ>

truncations. These appear to be localized uniformly throughout the cytoplasm suggesting that

differential cellular localisation cannot explain their apparent differential effects on cleavage

of various γ-secretase substrates. While the truncations appear to be excluded from the

nucleus, they show no obvious localisation to any particular cytosolic compartment although

no fluorescence was detected in lysosomes. There are a number of possible explanations for

the lack of expression in lysosomes. Firstly, EGFP is sensitive to low pH [37], therefore in

acidic environments such as lysosomes its fluorescence is quenched. It is possible that the

74

EGFP-Psen1Δ> truncation proteins are present within lysosomes but cannot be visualized

under fluorescence using GFP. Alternatively these proteins may not be accumulated into

lysosomes. A third possibility is that the turnover for these proteins within lysosomes is rapid.

Further work using a fluorescent protein with less sensitivity to low pH or the use of drugs

that raise lysosome pH are possible future tests to clarify whether or not the EGFP-Psen1Δ>

truncations are found within lysosomes. Alternatively, the use of polyclonal anti-GFP

antibodies to detect the GFP-fusion protein by immunocytochemistry can also be performed.

Although unlikely, the lack of fluorescence in lysosomes may be indicative of either a lack of

fusion between autophagosomes and lysosomes, or a reduced induction of autophagy, with

the truncated forms of Psen1 as a causal factor. It is known that PSEN1 is involved in

lysosomal acidification [38], and that the initial site of formation of autophagosomes is at

MAM contact sites in mammalian cells [39]. The MAM is highly enriched in PSENs [34].

Therefore it is necessary to further investigate the localization of the EGFP-Psen1Δ>

truncations within lysosomes.

Expression of the EGFP–Psen1 full length fusion protein could not be detected at 2.5 hpf as it

is difficult to force ectopic expression of full-length Presenilin protein in zebrafish embryos if

endogenous protein expression is still present. Therefore it is necessary to block endogenous

expression in order to visualise the localization of EGFP–Psen1. One possible way to block

endogenous protein expression would be to use morpholino antisense oligonucleotides that

specifically targets endogenous mRNA encoding for Psen1 to prevent translation by means of

a steric block mechanism.

75

Fluorescent tags are a useful way to visualize protein localization within a cell. However,

fusion of a sizeable protein such as EGFP to a protein of interest has the potential to disrupt

native protein function and localization [40]. Therefore, we cannot rule out the possibility that

the subcellular localization of the Psen1 truncations has been affected by the presence of

EGFP at their N-termini. One method to validate their supposed lack of localization would be

to create constructs encoding EGFP fused at the C-termini of the Psen1 truncations to

determine if these show a similar localization pattern to the N-terminal fusions. However for

multiple membrane-spanning proteins such as Psen1, fusion of EGFP at the C-terminus might

disrupt protein conformation within the membrane layer in some cases. Fusion of a much

smaller polypeptide protein tag such as the FLAG tag [41] to the protein of interest followed

by immunocytochemistry could be used as another method of confirmation of our in vivo

results.

76

Acknowledgements

This research was supported by the award to AL of an Adelaide Graduate Research

Scholarship from the University of Adelaide, by a grant from the Australian National Health

and Medical Research Council [NHMRC, Project Grant APP1061006 to ML and MN

http://www.nhmrc.gov.au/] and by funds from the School of Biological Sciences and

Department of Genetics and Evolution of the University of Adelaide [support of ML and

MN]. We are grateful for the support to MN given by the family of Lindsay Carthew.

Author Disclosure Statement

No competing financial interests exist.

77

Figure legends

Figure 1. The modified polymerase chain reaction (PCR) driven overlap extension

method. Template cDNA for EGFP (straight lines) and psen1 (dotted lines) was used in

initial PCRs to amplify the two fragments. Blue arrows indicate the EGFP primers a and b,

and the psen1 primers c and d. The reverse primer for EGFP is complementary at its 5' end to

the 5' end of the forward primer for zebrafish psen1. The two amplified fragments are fused

together in a second PCR in which only primers a and d are present. The chimeric cDNA is

then restricted and ligated into the vector pcGblobin2 after which mRNA can be synthesized

for injection into zebrafish embryos.

78

Figure 2. Fluorescent protein expression and dye staining following cell dissociation. a)

Live staining of cells with MitoTracker Red dye and the nuclear dye Hoechst 33342 (in blue).

The fluorescence intensity of the cells was not uniform, as can be seen for the cell on the left

indicated by an arrowhead. b) Expression of fluorescently-tagged nuclear H2B-RFP (in red)

and CAAX motif membrane-GFP (in green) proteins in cells. The cell to the right indicated

by an arrowhead has lower fluorescence intensity compared to adjacent cells.

79

Figure 3. Structure of the zebrafish Psen1 protein and truncations. The upper portion of

the figure shows the relationship of psen1 exons to protein structure. Alternating thick and

thin areas on the protein strand indicate the sequence coded by successive exons, and are

numbered after their cognate human exons. Horizontal grey lines represent the lipid bilayer.

‘D’ indicates the position of the two highly conserved aspartate residues critical for the

catalytic site (shown by a star). The arrowhead indicates the site of endoproteolysis that

creates the N- and C-terminal fragments and the N- and C-termini are indicated. The

cytoplasmic loop domain and the approximate locations of the G183V mutation (Pick disease)

and the P242LfsX11 mutation (Acne Inversa) in human PSEN1 are indicated. The lower

portion of the figure shows the putative structure of each truncation of Psen1 and their

observed effects on Notch signalling and Appa cleavage. Δ> denotes truncation after a

particular exon. ↑ indicates increase, ↓ indicates decrease, 0 indicates no significant change

(Reprinted by permission from Newman et al. (2013) [11])

80

81

Figure 4. Expression of the EGFP-Psen1 truncations in dissociated cells. Cells expressing

a) EGFP-Psen1Δ>4, b) EGFP-Psen1Δ>5, c) EGFP-Psen1Δ>6, d) EGFP-Psen1Δ>7, e) EGFP-

Psen1Δ>8. Embryos were stained with MitoTracker Red prior to cell dissociation.

Arrowheads indicate areas within the cell with no detectable EGFP fluorescence.

82

Figure 5. Regions within cells lacking EGFP fluorescence expression colocalize with

lysosomal dye stain. Cells expressing a) EGFP-Psen1Δ>4, b) EGFP-Psen1Δ>6. Embryos

were stained with LysoTracker Red prior to cell dissociation. Arrowheads indicate areas

within the cell with no detectable EGFP fluorescence but which show noticeable LysoTracker

Red indicating the presence of lysosomes.

83

References

1. De Jonghe, C., et al., Aberrant splicing in the presenilin-1 intron 4 mutation causes presenile Alzheimer's disease by increased Abeta42 secretion. Human molecular genetics, 1999. 8(8): p. 1529-40.

2. Laudon, H., et al., A nine-transmembrane domain topology for presenilin 1. The Journal of biological chemistry, 2005. 280(42): p. 35352-60.

3. Mann, D.M., et al., Cases of Alzheimer's disease due to deletion of exon 9 of the presenilin-1 gene show an unusual but characteristic beta-amyloid pathology known as 'cotton wool' plaques. Neuropathology and applied neurobiology, 2001. 27(3): p. 189-96.

4. Fukumori, A., et al., Three-Amino Acid Spacing of Presenilin Endoproteolysis Suggests a General Stepwise Cleavage of gamma-Secretase-Mediated Intramembrane Proteolysis. Journal of Neuroscience, 2010. 30(23): p. 7853-7862.

5. Sjogren, M. and C. Andersen, Frontotemporal dementia--a brief review. Mech Ageing Dev, 2006. 127(2): p. 180-7.

6. Dermaut, B., et al., A novel presenilin 1 mutation associated with Pick's disease but not beta-amyloid plaques. Annals of neurology, 2004. 55(5): p. 617-26.

7. Watanabe, H., et al., Familial frontotemporal dementia-associated presenilin-1 c.548G>T mutation causes decreased mRNA expression and reduced presenilin function in knock-in mice. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2012. 32(15): p. 5085-96.

8. Manabe, T., et al., HMGA1a: sequence-specific RNA-binding factor causing sporadic Alzheimer's disease-linked exon skipping of presenilin-2 pre-mRNA. Genes to cells : devoted to molecular & cellular mechanisms, 2007. 12(10): p. 1179-91.

9. Moussavi Nik, S.H., M. Newman, and M. Lardelli, The response of HMGA1 to changes in oxygen availability is evolutionarily conserved. Experimental cell research, 2011. 317(11): p. 1503-12.

10. Wang, B., et al., Gamma-secretase gene mutations in familial acne inversa. Science, 2010. 330(6007): p. 1065.

11. Newman, M., et al., Differential, dominant activation and inhibition of Notch signalling and APP cleavage by truncations of PSEN1 in human disease. Human molecular genetics, 2013.

12. Sato, N., et al., Increased production of beta-amyloid and vulnerability to endoplasmic reticulum stress by an aberrant spliced form of presenilin 2. The Journal of biological chemistry, 2001. 276(3): p. 2108-14.

13. Shaner, N.C., P.A. Steinbach, and R.Y. Tsien, A guide to choosing fluorescent proteins. Nature methods, 2005. 2(12): p. 905-9.

14. Prasher, D.C., et al., Primary structure of the Aequorea victoria green-fluorescent protein. Gene, 1992. 111(2): p. 229-33.

15. Gerdes, H.H. and C. Kaether, Green fluorescent protein: applications in cell biology. FEBS letters, 1996. 389(1): p. 44-7.

16. Contento, A.L., Y. Xiong, and D.C. Bassham, Visualization of autophagy in Arabidopsis using the fluorescent dye monodansylcadaverine and a GFP-AtATG8e fusion protein. Plant J, 2005. 42(4): p. 598-608.

17. Spandl, J., et al., Live Cell Multicolor Imaging of Lipid Droplets with a New Dye, LD540. Traffic, 2009. 10(11): p. 1579-1584.

18. Bereiter-Hahn, J., Dimethylaminostyrylmethylpyridiniumiodine (DASPMI) as a fluorescent probe for mitochondria in situ. Biochimica et Biophysica Acta (BBA) - Bioenergetics, 1976. 423(1): p. 1-14.

19. Draper, B.W., P.A. Morcos, and C.B. Kimmel, Inhibition of zebrafish fgf8 pre-mRNA splicing with morpholino oligos: a quantifiable method for gene knockdown. Genesis, 2001. 30(3): p. 154-6.

84

20. Ro, H., et al., Novel vector systems optimized for injecting in vitro-synthesized mRNA into zebrafish embryos. Mol Cells, 2004. 17(2): p. 373-6.

21. Burdall, S.E., et al., Breast cancer cell lines: friend or foe? Breast Cancer Res, 2003. 5(2): p. 89-95.

22. Westerfield, M., ed. The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio). 4th ed. ed. 2000, Univ. of Oregon Press, Eugene.

23. Heckman, K.L. and L.R. Pease, Gene splicing and mutagenesis by PCR-driven overlap extension. Nature protocols, 2007. 2(4): p. 924-32.

24. Hancock, J.F., et al., A CAAX or a CAAL motif and a second signal are sufficient for plasma membrane targeting of ras proteins. The EMBO journal, 1991. 10(13): p. 4033-4039.

25. Kanda, T., K.F. Sullivan, and G.M. Wahl, Histone–GFP fusion protein enables sensitive analysis of chromosome dynamics in living mammalian cells. Current Biology, 1998. 8(7): p. 377-385.

26. Bhasin, M., E.L. Reinherz, and P.A. Reche, Recognition and classification of histones using support vector machine. J Comput Biol, 2006. 13(1): p. 102-12.

27. Alexandre, P., et al., Neurons derive from the more apical daughter in asymmetric divisions in the zebrafish neural tube. Nature neuroscience, 2010. 13(6): p. 673-9.

28. Andersen, S.L., Preparation of dissociated Zebrafish spinal neuron cultures. Methods in Cell Science, 2001. 23(4): p. 205-209.

29. Annaert, W.G., et al., Presenilin 1 controls gamma-secretase processing of amyloid precursor protein in pre-golgi compartments of hippocampal neurons. The Journal of cell biology, 1999. 147(2): p. 277-94.

30. Pasternak, S.H., Presenilin-1, Nicastrin, Amyloid Precursor Protein, and -Secretase Activity Are Co-localized in the Lysosomal Membrane. Journal of Biological Chemistry, 2003. 278(29): p. 26687-26694.

31. Ankarcrona, M. and K. Hultenby, Presenilin-1 is located in rat mitochondria. Biochemical and biophysical research communications, 2002. 295(3): p. 766-70.

32. Li, J., et al., Alzheimer presenilins in the nuclear membrane, interphase kinetochores, and centrosomes suggest a role in chromosome segregation. Cell, 1997. 90(5): p. 917-27.

33. Rechards, M., et al., Presenilin-1 exists in both pre- and post-Golgi compartments and recycles via COPI-coated membranes. Traffic, 2003. 4(8): p. 553-65.

34. Area-Gomez, E., et al., Presenilins are enriched in endoplasmic reticulum membranes associated with mitochondria. The American journal of pathology, 2009. 175(5): p. 1810-6.

35. Snapp, E., Design and Use of Fluorescent Fusion Proteins in Cell Biology. Current protocols in cell biology / editorial board, Juan S. Bonifacino ... [et al.], 2005. CHAPTER: p. Unit-21.4.

36. Ludtmann, M.H., et al., An ancestral non-proteolytic role for presenilin proteins in multicellular development of the social amoeba Dictyostelium discoideum. Journal of cell science, 2014. 127(Pt 7): p. 1576-84.

37. Tsien, R.Y., The green fluorescent protein. Annu Rev Biochem, 1998. 67: p. 509-44. 38. Lee, J.H., et al., Lysosomal proteolysis and autophagy require presenilin 1 and are disrupted

by Alzheimer-related PS1 mutations. Cell, 2010. 141(7): p. 1146-58. 39. Hamasaki, M., et al., Autophagosomes form at ER-mitochondria contact sites. Nature, 2013.

495(7441): p. 389-93. 40. Skube, S.B., J.M. Chaverri, and H.V. Goodson, Effect of GFP tags on the localization of EB1 and

EB1 fragments in vivo. Cytoskeleton, 2010. 67(1): p. 1-12. 41. Hopp, T.P., et al., A Short Polypeptide Marker Sequence Useful for Recombinant Protein

Identification and Purification. Nat Biotech, 1988. 6(10): p. 1204-1210.

85

CHAPTER 4

Research Paper 3

Mitochondrion-ER apposition lengths in zebrafish

embryos under inhibition of presenilin1 and presenilin2

gene expression

86

STATEMENT OF AUTHOURSHIP

Mitochondrion-ER apposition lengths in zebrafish embryos under inhibition of

presenilin1 and presenilin2 gene expression.

Unsubmitted manuscript

Anne Hwee Ling Lim (Candidate/First author)

• Performed all experiments (except where noted) and wrote manuscript.

Certification that the statement of contribution is accurate

Signed

Morgan Newman (co-author)

• Performed all morpholino injections into embryos.

Signed

Michael Lardelli (co-author)

• Planned the research, supervised development of work and edited the manuscript.

Signed

Certification that the statement of contribution is accurate

Certification that the statement of contribution is accurate

87

Mitochondrion-ER apposition lengths in zebrafish embryos under

inhibition of presenilin1 and presenilin2 gene expression.

Anne Lim1*, Morgan Newman1, Michael Lardelli1.

1. Alzheimer’s Disease Genetics Laboratory, Department of Genetics and Evolution, School

of Biological Sciences, The University of Adelaide, South Australia, Australia.

Email: anne.lim@adelaide.edu.au*

morgan.newman@adelaide.edu.au

michael.lardelli@adelaide.edu.au

* Corresponding author: Anne Lim.

88

Abstract

• Background: Mutations in the genes PRESENILIN1 (PSEN1), PRESENILIN2 (PSEN2)

and AMYLOID BETA (A4) PRECURSOR PROTEIN (APP) have been identified in

familial Alzheimer’s disease (FAD), which makes up approximately 5% of total AD

cases. Some subcellular compartments of the cell appear capable of interacting with each

other physically and also chemically. For instance, mitochondria and endoplasmic

reticulum (ER) are able to form physical appositions by linking via tethers. The length of

mitochondrion-ER appositions is increased in Psen1-/-/Psen2-/- double knockout murine

embryonic fibroblasts and in AD patient fibroblasts. To determine the effect of inhibition

of zebrafish psen1 and psen2 activity on mitochondrion-ER appositions in the zebrafish

animal model, we injected zebrafish embryos with morpholinos (MOs) that inhibit

expression of zebrafish psen1 and psen2. We then analyzed mitochondrion-ER apposition

in neural cells under electron microscopy.

• Results: Our analysis showed no significant difference in mitochondrion-ER apposition

lengths between control MO and psen1 & psen2 MO co-injected embryos at 24 hours post

fertilization. Instead, the distribution of mitochondrion-ER apposition lengths into

different length classes was close to identical.

• Conclusions: While our observations differ from those of the murine and human studies,

this may be due to differences in cell type, cell age and differences in organism species.

Keywords

Zebrafish, morpholino, TEM, mitochondrion-ER apposition lengths.

89

Background

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by the occurrence of

memory loss in its initial stages, with other effects such as the impairment of speech and

motor ability, depression, hallucinations, behaviour disturbances and, ultimately death in

more advanced stages of the disease (reviewed in [1]). The major neuropathological hallmarks

of the disease in the brain are amyloid plaques that consist of mainly the Amyloid beta (Aβ)

peptides, and neurofibrillary tangles, which are hyperphosphorylated forms of the tau protein.

The exact mechanism of the disease remains unclear. There have been numerous hypotheses

suggested with the most widely accepted being the amyloid cascade hypothesis. This states

that the accumulation of Aβ either via overproduction or lack of clearance, leads to its

oligomerization and deposition in the brain and ultimately to neuronal dysfunction,

degeneration and death. However, the amyloid hypothesis does not cover nor explain other

features observed in AD patients such as aberrant calcium homeostasis (reviewed in [2]),

mitochondrial dysfunction (reviewed in [3]), increased levels of cholesterol [4, 5] and altered

fatty acid and phospholipid metabolism [6].

The majority of AD cases are sporadic (SAD), with a small number of cases that are familial

(FAD). FAD characteristically has an early age of onset (<65 years). Although only

accounting for a small percentage of AD cases, most of our understanding of the molecular

events underlying the development of Alzheimer’s comes from FAD since genetic analysis

can be used to identify the genes and proteins involved.

Mutations in the PRESENILIN genes (PSEN1 and PSEN2) [7, 8] and the AMYLOID BETA

(A4) PRECURSOR PROTEIN (APP) [9] gene have been identified in FAD. The

PRESENILIN proteins make up one of the components of the γ-secretase enzyme complex

90

which cleaves type I transmembrane proteins. APP is a substrate of γ-secretase and its

cleavage forms the Aβ peptides found in plaques [10]. PRESENILINS also function

independently of γ-secretase. For instance, the PRESENILIN holoprotein plays a role in

lysosomal-dependent proteolysis by interacting with the v-ATPase V0a1 subunit resulting in

the maintenance of autolysosome/lysosome acidification [11]. PRESENILINS also regulate

β-catenin stability via a pathway for phosphorylation of β-catenin independent of Axin [12]

and form Ca2+ leak channels in the endoplasmic reticulum that allow the release of Ca2+ to the

cytoplasm [13]. SAD accounts for the majority of Alzheimer’s disease cases (95%), and

usually shows late-onset (>65 years). A combination of genetic and environmental risk factors

[e.g. inheritance of one or two APOE ε4 alleles, high cholesterol levels, obesity and diabetes]

have been linked to SAD (reviewed in [14-16]). Although SAD and FAD show similar AD

pathology, different mechanisms may underlie them since genes involved in FAD are not

detected in genome-wide association studies of SAD [17].

The subcellular localizations of the components of γ-secretase and its substrate APP have

been of great interest because determining where these components reside within cells would

provide further insight into the pathogenesis of AD. Various studies have found

PRESENILINS located in almost all membranous compartments of the cell [18-24]. A recent

discovery by Area-Gomez and colleagues appears to have identified a compartment where

PSEN1 and PSEN2 are highly enriched. Using mammalian cells they found PRESENILINS

to be located predominantly in the endoplasmic reticulum (ER) and specifically in a sub-

compartment of the ER known as the mitochondria-associated ER membrane (MAM) [25].

The MAM is a lipid raft-like compartment [26] that contains various enzymes involved in key

functions including the synthesis and transfer of phospholipids [27], oxidative protein folding

(reviewed in [28]) and calcium homeostasis [29]. Interestingly, the MAM is also the site of

91

formation of autophagosomes [30]. The MAM is physically linked to the outer mitochondrial

membrane by protein tethers that are sufficiently stable for MAM to be co-isolated with

mitochondria by subcellular fractionation [27](reviewed in [31]). Further work by Area-

Gomez et al. looked at the degree of mitochondrion-ER apposition in double knockout murine

embryonic fibroblasts (MEFs) and AD patient cells by electron microscopy [26].

Interestingly, they observed that the length of mitochondrion-ER appositions is increased in

Psen1-/-/Psen2-/- double knockout MEFs and also in both FAD and SAD patient fibroblasts

compared to controls [26]. Increased levels of phospholipid and cholesteryl ester synthesis in

MEFs and AD patient fibroblasts were also seen. This suggests that increased mitochondrial-

MAM communication may be a common factor in familial and sporadic AD. Perturbations to

MAM function might possibly result in various features of AD that the amyloid hypothesis

fails to explain.

The zebrafish is an advantageous model organism for the study of human disease. Zebrafish

embryos are robust and can undergo experimental manipulations such as the injection of

morpholino oligonucleotides into embryos to modify simultaneously the expression of

multiple genes. Zebrafish embryos are produced in large numbers and develop rapidly [32].

Zebrafish possess genes orthologous to human APP, PSEN1 and PSEN2 along with the other

components of the γ-secretase complex [33-35]. To date, the zebrafish animal model has not

been used to study mitochondrion-ER apposition. Our previous work has revealed aspects of

Presenilin protein function not seen in cell culture systems [36-38]. We were, therefore,

curious to see whether zebrafish might be an advantageous model in which to study the

effects of Presenilin activity on mitochondrion-ER apposition. Here, we determine the effect

of inhibition of psen1 and psen2 mRNA translation on mitochondrion-ER apposition in

92

midline spinal cord cells in the trunk region of 24 hours post fertilization (hpf) zebrafish

embryos.

Methods

Ethics statement

All zebrafish work was performed under the auspices of The Animal Ethics Committee of

The University of Adelaide

Animal husbandry

Wild-type zebrafish were maintained in a recirculated water system with a 14 hour light/10

hour dark cycle. Fertilized embryos were grown at 28˚C in embryo medium (E3)[39].

Morpholino microinjection of zebrafish embryos

Morpholinos were synthesized by Gene Tools LLC (Corvallis, OR, USA) and are listed in

Table 1. Fertilized zebrafish embryos were rinsed in E3 medium and injected at the one cell

stage. These morpholino-injected embryos were grown to 24 hpf before fixation was

performed. To ensure consistency of morpholino injections, zebrafish eggs were always

injected with solutions at 1 mM total concentration (i.e. 0.5 mM MoPsen1Tln + 0.5 mM

MoPsen2Tln, or 1.0 mM MoCont alone).

Transmission electron microscopy

24 hpf zebrafish larvae were fixed in 4% paraformaldehyde, 1.25% glutaraldehyde in PBS

with 4% sucrose, pH 7.2 overnight at 4˚C, and then rinsed in washing buffer (PBS with 4%

sucrose). Embryos were post-fixed in 2% osmium tetroxide, followed by dehydration through

an ethanol series, and rinsed with propylene oxide followed by resin infiltration. Embedded

93

embryos were sectioned transversely and posterior to the embryonic yolk ball, at the

beginning of the yolk extension using an ultramicrotome to obtain several 85nm thick

sections of the spinal cord. The ultrathin sections were stained with uranyl acetate and lead

citrate and imaged on an Olympus-SIS Veleta CCD camera in a FEI Tecnai G2 Spirit TEM.

Images were obtained from 3 cells at the midline area of the spinal cord. This was performed

for 3 embryos from each treatment. The mitochondrion-ER apposition lengths were measured

using Image J software and analysed statistically using Fisher’s exact test.

Results and Discussion

A recent study by Area-Gomez et al. found that MEFs as well as fibroblasts from patients

with FAD and SAD show increased length of mitochondrion-ER appositions compared to

wild type (WT) cells [26]. They found relatively more appositions in the 50-200nm (long) and

>200nm (very long) ranges compared to WT MEFs which have relatively more appositions

with a shorter length of <50nm. Zebrafish have proven to be a highly manipulable model for

analysis of PRESENILIN gene function and might prove to be a useful tool for analysis of the

role of these genes in communication between the ER and mitochondria. Therefore, we

sought to determine the effect of inhibition of psen1 and psen2 activity on mitochondrion-ER

apposition in the zebrafish animal model. To do this we simultaneously blocked translation of

the Presenilin proteins by injecting zebrafish embryos at the 1-cell stage with morpholinos

binding over the start codons of the psen1 and psen2 mRNAs (MoPS1Tln and MoPS2Tln

respectively). We injected an inactive morpholino as a negative control (MoCont). To

standardise the area of the embryo being observed, we chose to section 24 hpf embryos

transversely and posterior to the embryonic yolk ball, at the beginning of the yolk extension

(i.e. in the trunk region, Figure. 1). We chose to examine mitochondrion-ER apposition

lengths in neural cells at the midline of the developing spinal cord under the assumption that

94

these are most likely to represent neural precursor cells and therefore be a reasonably

homogenous cell type (Figure. 2). Three embryos injected with MoCont and three embryos

injected with both MoPS1Tln and MoPS2Tln were analysed. Mitochondrion-ER apposition

lengths were examined in three midline cells from each embryo (Figure 3). A total of 23

appositions in MoCont-injected embryos and 25 appositions in MoPS1Tln and MoPS2Tln co-

injected embryos were observed.

We observed that the majority of mitochondrion-ER apposition lengths for MoPS1Tln and

MoPS2Tln co-injected 24hpf embryos were long followed by short appositions and very long

appositions at 68%, 28% and 4% respectively. However, this distribution of apposition

lengths was also seen in the MoCont-injected embryos, i.e. 65.2% long, 30.4% short and

4.4% very long appositions (Figure. 4a and b and Table 2.1 and 2.2). We further divided the

long apposition lengths into smaller interval class lengths of 50-100nm, 100-150nm and 150-

200nm to determine if changes in apposition were occurring along a smaller length range.

Analysis by Fisher’s exact test showed no significant difference in apposition length

distribution between control embryos and those lacking Presenilin activity. Therefore the

distribution of mitochondrion-ER apposition lengths into different length classes at 24 hpf is

similar between embryos injected with negative control morpholino and embryos co-injected

with MoPS1Tln and MoPS2Tln.

Our observations do not appear consistent with those of the 2012 study by Area-Gomez et al..

There are several possible explanations for this. Firstly, the ability of the MoPS1Tln plus

MoPS2Tln morpholino injection to suppress expression of the psen1 and psen2 genes may be

questioned. A known phenotypic effect of the loss of Psen1 and Psen2 activity in zebrafish

embryos at 48 hpf is a decrease in the size and number of melanocytes such that surface

95

pigmentation overall appears reduced [36]. To check the effectiveness of our injection of

morpholinos to affect psen1 and psen2 activity, we injected embryos with the same batch of

morpholinos used in our TEM study above and confirmed the previously seen phenotypic

effects at 48 hpf. MoPS1Tln plus MoPS2Tln injected embryos showed reduced pigmentation

while MoCont-injected embryos were indistinguishable from wild type uninjected controls

(data not shown). Based on this result, we believe that significant loss of Presenilin activity

was achieved in 24 hpf embryos, and the efficiency of the morpholinos in inhibition of

translation was unlikely to cause the differences in mitochondrion-ER apposition lengths

observed between mammalian cells and zebrafish. However, it is possible that sufficient

zebrafish Presenilin protein remains at 24 hours to allow normal mitochondrion-ER

apposition. Future studies should control more closely for Presenilin protein loss at 24 hpf

after morpholino injection. One technique would be the use of immunohistochemistry on

tissue sections to confirm morpholino efficiency at inhibiting Presenilin protein translation. A

second possibility concerns the age of the embryos. Changes in apposition length may require

a longer time frame than 24 hours of embryonic development. It may be that a transient

knockdown of Psen1 and Psen2 does not permit sufficient time for noticeable changes in

mitochondrion-ER apposition lengths to occur when compared to controls. In mammalian

cells, the majority of apposition lengths in wild type cells were punctate (< 50 nm) while in

our study, the majority apposition lengths fall into the 50-200 nm range. The cell type

examined may also be a cause of this difference. Area-Gomez et al. [26] used fibroblast cells

from mice and AD patients while we examined putative neuronal progenitor cells that

probably have quite different energy and substrate requirements relative to fibroblasts. These

differences may affect the interaction between mitochondria and the ER. The psen1 and psen2

mRNA can be detected ubiquitously at early stages of embryogenesis, well before 24 hpf [34,

35, 40]. However the possibility that the putative neuronal progenitor cells examined do not

96

express Presenilin proteins during the timepoint analyzed must be recognized. Assaying for

the presence of the proteins in these cell types using immunohistochemistry will clarify this

issue. Lastly, although the number of total contacts analyzed for MoCont and MoPS1Tln plus

MoPS2Tln injected embryos were comparable to the number of contacts analysed by Area-

Gomez et al. in AD patient fibroblasts, it cannot be excluded that mosaicism in knockdown

due to an uneven distribution of morpholinos during injection of embryos could affect the

results obtained from the small sample size of 3 cells analysed per embryo. Such mosaicism is

presumably small in the AD patient fibroblasts used by Area-Gomez et al.. Therefore

increasing the number of cells analyzed will be necessary to reduce the margin of error and

increase accuracy.

Conclusions

We observed no significant difference in the distribution of mitochondrion-ER apposition

lengths between embryos injected with negative control morpholino and those injected with

morpholinos for simultaneous blockage of psen1 and psen2 expression at 24 hours post

fertilization. The difference between our observations and those from studies performed in

mammalian systems may be due to differences in cell type, cell age and/or differences in

organism species. Future studies should examine other zebrafish cell types and ages and

should control closely for loss of Presenilin protein expression.

97

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

AL carried out the TEM imaging and analysis, and drafted the manuscript. MN carried out the

morpholino injections of embryos and contributed to the design of the study. ML conceived

of the study, partook in its design and helped draft the manuscript. All authors read and

approved the final manuscript.

Acknowledgements

We would like to acknowledge Ruth Williams and Lyn Waterhouse from Adelaide

Microscopy for their technical assistance in TEM. This work was supported by funds from

The School of Molecular and Biomedical Sciences of the University of Adelaide and from

NHMRC project grant APP1061006 to MN and ML.

98

Tables

Table 1. Morpholino sequences

Morpholino Morpholino Sequences (5ˊ - 3ˊ) MoCont MoPS1Tln MoPS2Tln

CCTCTTACCTCAGTTACAATTTATA ACTAAATCAGCCATCGGAACTGTGA GTGACTGAATTTACATGAAGGATGA

Table 2.1. Percentage observed per length class, 3 classes.

Treatment Length of region of apposition, mitochondrial perimeter (nm) <50nm 50-200nm >200nm

MoCont (n=23) 30.4% (7) 65.2% (15) 4.4% (1) MoPS1Tln/MoPS2Tln injected (n=25)

28% (7) 68% (17) 4% (1)

Table 2.2. Percentage observed per length class, 5 classes.

Treatment Length of region of apposition, mitochondrial perimeter (nm)

<50nm 50-100nm >100-150nm

>150-200nm >200nm

MoCont (n=23) 30.4% (7) 34.8% (8) 17.4% (4) 13% (3) 4.4% (1) MoPS1Tln/MoPS2Tln injected (n=25)

28% (7) 44% (11) 8% (2) 16% (4) 4% (1)

99

Figure Legends

Figure 1. Zebrafish embryo at 24 hpf. Several sections of 85 nm thickness were obtained

from the area posterior to the yolk ball, within the yolk extension as indicated by the vertical

line.

100

Figure 2. The midline of the spinal cord region. Mitochondrion-ER apposition lengths were

measured in 3 cells at the midline area of the spinal cord in each embryo. Three MoCont-

injected embryos and three MoPS1Tln/MoPS2Tln-injected embryos were examined. The

midline of the spinal cord is indicated with a dashed line. Typical cells in which

mitochondrion-ER apposition lengths would be examined are indicated with arrows.

101

Figure 3. Electron microscopy of zebrafish neural cells. Cells were treated with

morpholinos binding the start codon of psen1 and psen2 mRNA (MoPS1Tln and MoPS2Tln

respectively). Arrowheads indicate the region of apposition between mitochondria (M) and

endoplasmic reticulum (E) i.e. MAM.

102

Figure 4. Quantitation of mitochondrion-ER apposition lengths (length of region of

contact). A) 3 length classes. B) 5 length classes. Total apposition events analysed was 23 in

MoCont and 25 in MoPS1Tln/MoPS2Tln-injected embryos, p = 1.00 for A and p = 0.889 for

B).

103

References

1. Voisin, T. and B. Vellas, Diagnosis and treatment of patients with severe Alzheimer's disease. Drugs & aging, 2009. 26(2): p. 135-44.

2. Bezprozvanny, I. and M.P. Mattson, Neuronal calcium mishandling and the pathogenesis of Alzheimer's disease. Trends in neurosciences, 2008. 31(9): p. 454-63.

3. Wang, X., et al., The role of abnormal mitochondrial dynamics in the pathogenesis of Alzheimer's disease. Journal of neurochemistry, 2009. 109 Suppl 1: p. 153-9.

4. Kuo, Y.M., et al., Elevated low-density lipoprotein in Alzheimer's disease correlates with brain abeta 1-42 levels. Biochemical and biophysical research communications, 1998. 252(3): p. 711-5.

5. Pappolla, M.A., et al., Mild hypercholesterolemia is an early risk factor for the development of Alzheimer amyloid pathology. Neurology, 2003. 61(2): p. 199-205.

6. Pettegrew, J.W., et al., Brain membrane phospholipid alterations in Alzheimer's disease. Neurochemical research, 2001. 26(7): p. 771-82.

7. Sherrington, R., et al., Cloning of a gene bearing missense mutations in early-onset familial Alzheimer's disease. Nature, 1995. 375(6534): p. 754-60.

8. Levy-Lahad, E., et al., Candidate gene for the chromosome 1 familial Alzheimer's disease locus. Science, 1995. 269(5226): p. 973-7.

9. Goate, A., et al., Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature, 1991. 349(6311): p. 704-6.

10. De Strooper, B., et al., Deficiency of presenilin-1 inhibits the normal cleavage of amyloid precursor protein. Nature, 1998. 391(6665): p. 387-90.

11. Lee, J.H., et al., Lysosomal proteolysis and autophagy require presenilin 1 and are disrupted by Alzheimer-related PS1 mutations. Cell, 2010. 141(7): p. 1146-58.

12. Kang, D.E., et al., Presenilin couples the paired phosphorylation of beta-catenin independent of axin: implications for beta-catenin activation in tumorigenesis. Cell, 2002. 110(6): p. 751-62.

13. Tu, H., et al., Presenilins form ER Ca2+ leak channels, a function disrupted by familial Alzheimer's disease-linked mutations. Cell, 2006. 126(5): p. 981-93.

14. Martins, I.J., et al., Apolipoprotein E, cholesterol metabolism, diabetes, and the convergence of risk factors for Alzheimer's disease and cardiovascular disease. Molecular psychiatry, 2006. 11(8): p. 721-36.

15. Gustafson, D., et al., An 18-year follow-up of overweight and risk of Alzheimer disease. Archives of internal medicine, 2003. 163(13): p. 1524-8.

16. Luchsinger, J.A., et al., Diabetes mellitus and risk of Alzheimer's disease and dementia with stroke in a multiethnic cohort. American journal of epidemiology, 2001. 154(7): p. 635-41.

17. Bertram, L. and R.E. Tanzi, Genome-wide association studies in Alzheimer's disease. Human molecular genetics, 2009. 18(R2): p. R137-45.

104

18. Annaert, W.G., et al., Presenilin 1 controls gamma-secretase processing of amyloid precursor protein in pre-golgi compartments of hippocampal neurons. The Journal of cell biology, 1999. 147(2): p. 277-94.

19. Kimura, N., et al., Age-related changes in the localization of presenilin-1 in cynomolgus monkey brain. Brain research, 2001. 922(1): p. 30-41.

20. Vetrivel, K.S., et al., Association of gamma-secretase with lipid rafts in post-Golgi and endosome membranes. The Journal of biological chemistry, 2004. 279(43): p. 44945-54.

21. Pasternak, S.H., Presenilin-1, Nicastrin, Amyloid Precursor Protein, and -Secretase Activity Are Co-localized in the Lysosomal Membrane. Journal of Biological Chemistry, 2003. 278(29): p. 26687-26694.

22. Ankarcrona, M. and K. Hultenby, Presenilin-1 is located in rat mitochondria. Biochemical and biophysical research communications, 2002. 295(3): p. 766-70.

23. Li, J., et al., Alzheimer presenilins in the nuclear membrane, interphase kinetochores, and centrosomes suggest a role in chromosome segregation. Cell, 1997. 90(5): p. 917-27.

24. Marambaud, P., et al., A presenilin-1/gamma-secretase cleavage releases the E-cadherin intracellular domain and regulates disassembly of adherens junctions. The EMBO journal, 2002. 21(8): p. 1948-56.

25. Area-Gomez, E., et al., Presenilins are enriched in endoplasmic reticulum membranes associated with mitochondria. The American journal of pathology, 2009. 175(5): p. 1810-6.

26. Area-Gomez, E., et al., Upregulated function of mitochondria-associated ER membranes in Alzheimer disease. The EMBO journal, 2012. 31(21): p. 4106-23.

27. Vance, J.E., PHOSPHOLIPID-SYNTHESIS IN A MEMBRANE-FRACTION ASSOCIATED WITH MITOCHONDRIA. Journal of Biological Chemistry, 1990. 265(13): p. 7248-7256.

28. Raturi, A. and T. Simmen, Where the endoplasmic reticulum and the mitochondrion tie the knot: the mitochondria-associated membrane (MAM). Biochimica et biophysica acta, 2013. 1833(1): p. 213-24.

29. Csordas, G., et al., Imaging interorganelle contacts and local calcium dynamics at the ER-mitochondrial interface. Molecular cell, 2010. 39(1): p. 121-32.

30. Hamasaki, M., et al., Autophagosomes form at ER-mitochondria contact sites. Nature, 2013. 495(7441): p. 389-93.

31. Vance, J.E., MAM (mitochondria-associated membranes) in mammalian cells: Lipids and beyond. Biochimica et biophysica acta, 2013.

32. Dahm, R. and R. Geisler, Learning from small fry: the zebrafish as a genetic model organism for aquaculture fish species. Marine biotechnology, 2006. 8(4): p. 329-45.

33. Musa, A., H. Lehrach, and V. Russo, Distinct expression patterns of two zebrafish homologues of the human APP gene during embryonic development. Development genes and evolution, 2001. 211(11): p. 563-567.

34. Leimer, U., et al., Zebrafish (Danio rerio) presenilin promotes aberrant amyloid beta-peptide production and requires a critical aspartate residue for its function in amyloidogenesis. Biochemistry, 1999. 38(41): p. 13602-9.

35. Groth, C., et al., Identification of a second presenilin gene in zebrafish with similarity to the human Alzheimer's disease gene presenilin2. Development genes and evolution, 2002. 212(10): p. 486-90.

36. Nornes, S., et al., Independent and cooperative action of Psen2 with Psen1 in zebrafish embryos. Experimental cell research, 2009. 315(16): p. 2791-801.

105

37. Newman, M., et al., Altering presenilin gene activity in zebrafish embryos causes changes in expression of genes with potential involvement in Alzheimer's disease pathogenesis. Journal of Alzheimer's disease : JAD, 2009. 16(1): p. 133-47.

38. Nornes, S., et al., Interference with splicing of Presenilin transcripts has potent dominant negative effects on Presenilin activity. Human molecular genetics, 2008. 17(3): p. 402-12.

39. Westerfield, M., ed. The zebrafish book. A guide for the laboratory use of zebrafish (Danio rerio). 4th ed. ed. 2000, Univ. of Oregon Press, Eugene.

40. Rauch, G.J., et al. Submission and Curation of Gene Expression Data. ZFIN Direct Data Submission. 2003; Available from: (http://zfin.org).

106

CHAPTER 5

Research Paper 4

Assessment of proteins present in the mitochondria-

associated membrane, a sub-compartment of the

endoplasmic reticulum.

107

STATEMENT OF AUTHOURSHIP

Assessment of proteins present in the mitochondria-associated membrane, a sub-

compartment of the endoplasmic reticulum.

Unsubmitted manuscript

Anne Hwee Ling Lim (Candidate/First author)

• Performed all experiments and wrote manuscript.

Certification that the statement of contribution is accurate

Signed

Michael Lardelli (co-author)

• Planned the research, supervised development of work and edited the manuscript.

Certification that the statement of contribution is accurate

Signed

108

Assessment of proteins present in the mitochondria-associated membrane, a

sub-compartment of the endoplasmic reticulum.

Anne Lim1, Michael Lardelli1.

1. Alzheimer’s Disease Genetics Laboratory, Department of Genetics and Evolution, School

of Biological Sciences, The University of Adelaide, South Australia, Australia.

Corresponding author: Anne Lim

Tel.: +61 8 8313 4863

Fax: +61 8 8313 7534

Email: anne.lim@adelaide.edu.au

109

Abstract

The different functions of a cell are performed by various organelle compartments, with each

structure containing its own assortment of proteins. One such compartment is the

mitochondria associated membrane, a sub-compartment of the endoplasmic reticulum that

performs specialized functions in the cell. It is also one of the subcellular locations of the

components of the gamma secretase complex, an enzyme complex involved in the production

of the Aβ peptide, the main component of amyloid plaques in the Alzheimer’s disease brain.

Using Western immunoblot analysis, we studied the presence or absence of several proteins

of interest in four subcellular fractions. Among the proteins analysed, the SORTILIN-

RELATED RECEPTOR was of great interest as it has been identified in several genome-wide

association studies to associate with sporadic late onset Alzheimer’s disease. Our results show

that SORL1 is present in the MAM.

Introduction

Subcellular fractionation studies have found that mitochondria can be isolated with a sub-

compartment of the endoplasmic reticulum (ER) known as the mitochondria associated

membrane (MAM) [1]. The MAM is a specialized subdomain of the ER that is physically

tethered and biochemically connected to the mitochondria. The MAM has the characteristics

of an intracellular lipid raft [2], such as its enrichment in cholesterol and lipids [3, 4]. The

links between both organelles are formed from proteins such as phosphofurin acidic cluster

sorting protein 2 (PACS2) [5] and mitofusin-2 (MFN2) [6], allowing for intercommunication

to occur.

110

The MAM is involved in several key functions. One such function is its role in cholesterol

metabolism. The MAM is enriched in the enzyme acyl-coenzyme A (CoA):cholesterol

acyltransferase 1 (ACAT1) [7]. ACAT1 is involved in the conversion of free cholesterol to

cholesteryl esters which are stored mainly as lipid droplets [8]. Another role of the MAM is in

lipid metabolism. For instance, the MAM and mitochondria collaborate to synthesize

phosphatidylethanolamine (PE) [9, 10]. The MAM is also enriched in calcium related proteins

such as the sigma-1 receptors [11] and the inositol-1,4,5-trisphosphate (IP3) receptors [12],

and appears to regulate Ca2+ levels from the ER to the mitochondria [13]. Intriguingly, the

MAM appears to be the site of formation of autophagosomes, with studies showing that

autophagosomal marker proteins were localised to contact sites during induction of autophagy

[14]. Additionally, depletion of MFN2 results in impairment of autophagosome formation

[15]. The MAM also has protein-folding chaperones and oxidoreductases and is a site of

disulfide bond formation (reviewed in [16]).

More focus has been given to the MAM in recent years due to its potential link to

neurodegenerative diseases. For instance, the MAM is the predominant location of the

PRESENILINS (PSEN1 and PSEN2), proteins implicated in familial Alzheimer’s disease

(FAD) [17]. The PRESENILINS are the catalytic unit of the enzyme γ-secretase complex

which is involved in the production of Aβ, the peptide that predominantly forms amyloid

plaques in the brains of AD patients. The majority of FAD is due to mutations in the gene

encoding for the PSEN1 protein, PSEN1. The MAM also contains the three other major

components which make up the γ-secretase complex, as well as the γ-secretase substrate

AMYLOID BETA (A4) PRECURSOR PROTEIN (APP) from which the Aβ peptide is

derived. The α-synuclein protein which makes up the major constituent of Lewy bodies, the

pathological hallmark of Parkinson’s disease (PD), has also been localised to the MAM [18].

111

Here we determine via Western immunoblot analysis if the AD associated proteins BETA

APP CLEAVING ENZYME (BACE1) and SORTILIN-RELATED RECEPTOR (SORL1)

are present at the MAM. We also examine the localization of Aβ associated 37kDa/67kDa

Laminin Receptor.

Results and discussion:

Analysis of the presence of proteins involved in Aβ production or trafficking in MAM.

The MAM has been discovered as the predominant location of PSEN1, and also contains the

other three components of the γ-secretase complex [17]. We have confirmed the localization

of PSEN1 to the MAM in mouse brain tissues (Fig. 1, data also available as a supplemental

data file to [19], Appendix I). The γ-secretase substrate, APP also resides at the MAM [17],

therefore we wanted to look at other proteins potentially related to the production or

trafficking of Aβ. The discovery of γ-secretase activity in MAM [17] made us question the

localization of BACE. BACE is involved in the amyloidogenic pathway where it cleaves APP

to release soluble, secreted β-APPs and an embedded C99 fragment. This C99 fragment is

subsequently cleaved by γ-secretase to generate smaller Aβ peptides of different sizes, such as

Aβ40 and Aβ42.

Two homologues of BACE, BACE1 and BACE2 have been identified [20], both of which can

cleave APP [21]. BACE1 is highly expressed in the pancreas and brain but BACE2 is

expressed at very low levels in the brain and is mostly expressed in peripheral tissues

(reviewed in [22, 23]). In the cell, BACE1 has thus far been localized mainly to the trans-

112

Golgi network (TGN) and endosomes [24]. However, it is not known if BACE1 localizes to

the MAM.

We performed a Western immunoblot analysis using the mouse mitochondrial, ER, MAM and

plasma membrane (PM) fractions. The MAM and ER fractions were free of contamination.

However there was cross-contamination between the mitochondrial and PM fractions (Fig.

2d). Due to this, the results of the immunoblot analysis were focused mainly on the MAM and

ER fractions.

We analysed the fractions for the presence of BACE1. Our Western blot analysis indicated a

strong band that was the expected size of 70kDa in the mitochondria, ER and PM fraction

(Fig. 2a). It should be noted that Area-Gomez et al. have shown that the PM fractionation

protocol results in the Golgi fraction co-purifying with the PM fraction [17]. However no

clear band was visible in the MAM fraction.

Although there was cross-contamination, the presence of BACE1 in the PM/Golgi fraction

concurs with previous studies that BACE1 localises to the TGN. However the presence of

BACE1 in mitochondria has to our knowledge not been detected in previous studies and is

probably due to the cross-contamination with the PM/Golgi fraction (Fig. 2d). Its localization

in the ER is unsurprising as BACE1 travels from the ER to the TGN [25]. Unlike PSEN1, the

MAM is not the predominant location of BACE1 even though γ-secretase activity and APP

are present there. This could be a cellular mechanism by which the cell prevents an

overproduction of Aβ since cleavage of APP by BACE1 is the rate limiting step in the

113

amyloidogenic pathway (reviewed in [26]). Besides the MAM, APP has previously been

observed in endosomes, lysosomes [27], TGN, the late secretory pathway and at the cell

surface [24]. A recent study using neuronal cells indicates that, under basal conditions, the

majority of BACE1 localizes to recycling endosomes, with only 10-30% localizing at the

TGN, while the majority of APP resides in the Golgi therefore spatially separating substrate

and enzyme [25]. When neurons are stimulated with glycine to mimic induction of neuronal

activity in the brain, more APP can be observed in recycling endosomes [25]. The activity of

BACE1 has a low pH optimum [28]. Therefore it is likely that APP cleavage by BACE1

occurs in the acidic environment of endosomes and not likely at the MAM, thereby limiting

its presence in this compartment. The C99 fragment formed from the cleavage of APP by

BACE1 could be transported back to the MAM for further processing by the γ-secretase

complex, as the γ-secretase-cleaved APP intracellular domain (AICD) which is formed after

cleavage of C99 is detected at the MAM [17]. Further analysis of MAM fractions will look

for the presence of C99.

Next we examined a protein that is involved in trafficking of APP and possibly Aβ, SORL1.

SORL1 is a 250kDa transmembrane receptor that binds APP [29] and controls APP

trafficking through the TGN to endosomes. Multiple SORL1 alleles are associated with AD

risk [30]. SORL1 appears to be localized in the Golgi compartment, plasma membrane and

endosomes [29, 31]. We performed a Western blot analysis to determine if it could also be

localised in the MAM. Interestingly SORL1 protein is detected in the MAM (Fig. 2b). The

presence of both SORL1 and APP in the MAM may be an indication that the MAM is the

initial site of SORL1 receptor binding with APP, from which APP can then be trafficked to

the TGN, with regulated release from TGN to endosomes, lysosomes or plasma membrane.

114

Another receptor with interactions with Aβ is the 37kDa/67kDa Laminin Receptor (also

known as LRP/LR or RPSA). LRP/LR is a multifunctional receptor that acts as a cell surface

receptor for the cellular prion protein PrPc and as well as its misfolded form PrPSc. PrPc has

been shown to bind Aβ oligomers within cholesterol-rich lipid rafts upon which it is

internalized and trafficked to lysosomes, all of which requires the transmembrane low density

lipoprotein receptor-related protein-1 (LRP1) [32]. Interestingly, LRP/LR is also found in

lipid rafts and also interacts with Aβ resulting in its internalization at the cell surface [33].

Fluorescence studies indicate that LRP/LR localises to the cell surface, nucleus and to a

perinuclear compartment [34]. We hypothesized that this perinuclear compartment might be

the MAM. However, our Western blot analysis revealed no detectable LRP/LR within the

MAM fraction. Instead, LRP/LR was detected mainly in the ER fraction, which is likely the

perinuclear compartment (Fig. 2c).

Conclusion

We analysed three proteins for their presence in the MAM and have found that SORL1 can be

found in the MAM fraction. Its role in the MAM has yet to be elucidated. SORL1 is involved

in APP trafficking, with mutations linking it to AD. Understanding its role in MAM may give

insight into APP trafficking through the cell and may also clarify the role the MAM plays in

AD.

115

Methods and materials

Antibodies

Primary antibodies were against the proteins BACE1 (Santa Cruz Biotechnology, CA, USA,

Cat. No sc-10053), SORL1 (Aviva Systems Biology, CA, USA, Cat. No OAEB01612),

LRP/LR (Aviva Systems Biology, Cat. No ARP54683_P050), MTCO1 (Abcam, Cat. No.

ab14705), Na+/K+-ATPase (Abcam, Cat. No. ab7671), IP3R3 (Millipore, Temecula, USA,

Cat. No. AB9076), and KDEL (Abcam, ab12223). Secondary antibodies were donkey anti-

mouse Ig, (Rockland, Gilbertsville, USA, Cat. No. 610-703-124) and donkey anti-rabbit Ig,

(Rockland, 611-703-127).

Preparation of subcellular fractions

Mice brain tissue were quickly removed and placed in isolation buffer (250 mM mannitol, 5

mM HEPES pH 7.4, and 0.5 mM EGTA). The tissue was minced with scissors and gently

homogenized with 4-6 strokes in a loose Potter-Elvehjem grinder (Thermo Fisher, MA,

USA). The homogenate was centrifuged for 5 minutes at 1000 X g. This gave a pellet of cell

debris and nuclei which was discarded. The supernatant was centrifuged for 15 minutes at

10000 X g to pellet crude mitochondria which was separated from the supernatant. The

supernatant was centrifuged for 1 hour at 100000 X g in a Beckman Ti-70 rotor (Beckman

Coulter, CA, USA) to pellet the ER/microsomal fraction. The crude mitochondrial fraction

was layered on top of a 30% Percoll (GE Healthcare, UK) gradient and centrifuged for 30

minutes at 95000 X g in a Beckman Ti-70 rotor. The upper band contained the MAM fraction.

The lower band contained mitochondria fraction that was free of ER. The upper band was

diluted fivefold with isolation buffer and washed twice at 8000 X g for 10 minutes to obtain

the mitochondrial fraction in the pellet. The supernatant containing the MAM was centrifuged

116

at 100000 X g for 1 hour to obtain the MAM pellet which was resuspended in isolation buffer.

The lower band was diluted fivefold with isolation buffer and centrifuged twice at 8000 X g

for 10 minutes to remove any Percoll present. The resulting pellet was resuspended in

isolation buffer.

For the plasma membrane fraction, minced tissue was homogenised in STM 0.25 buffer (0.25

M sucrose, 10 mM Tris_Cl pH 7.4, 1.0 mM MgCl2; 4.5 ml/g tissue), using 10 strokes of the

loose-fitting Potter-Elvehjem grinder. The homogenate was centrifuged for 5 minutes 260 X

g. The supernatant was kept on ice while the pellet was resuspended in half the volume of

STM 0.25 buffer, homogenised again with three strokes of the loose gringer and centrifuged 5

minutes 260 X g. The two supernatants were combined, centrifuged for 10 minutes at 1500 X

g and the pellet resuspended in twice the volume of STM 0.25 buffer initially used. The

resuspended pellet was homogenised using three strokes of a tight fitting Potter-Elvehjem

grinder. An equal volume of STM 2 buffer (2 M sucrose, 10 mM Tris_Cl pH 7.4, 1.0 mM

MgCl2) was added to the homogenate, and centrifuged for 1 hour at 113,000 X g. The layer at

the top of the gradient is enriched in plasma membrane, and was resuspended in STM 0.25

buffer. All fractions were quantitated for total protein content using the Bradford system

(BioRad, CA, USA).

Western immunoblot assays for fractions

20 μg of total protein was placed in sample buffer [2% sodium dodecyl sulfate (SDS), 5% b-

mercaptoethanol, 25% v/v glycerol, 0.0625 M Tris–HCl (pH 6.8), and bromphenol blue],

heated at 55˚C for 10 min, and separated on 4-12% SDS–PAGE gel. Proteins were transferred

to nitrocellulose membranes using a semidry electrotransfer system. Western immunoblotting

used the conditions recommended by the antibody providers and visualized with luminol

117

reagents (Amresco, OH, USA) using the ChemiDocTM MPimaging system (Bio-Rad, CA,

USA).

118

Figures

Figure 1. The majority of PSEN1 is localized to the MAM in mammalian (mouse) brain

tissues. Fractionation of mouse brain cellular membranes was done to obtain plasma membrane

(PM), mitochondrial membranes (Mito), mitochondrial associated membranes (MAM) and non-

MAM endoplasmic reticulum (ER). Fractions were identified by their physical behaviour during

the fractionation procedure and confirmed by western immunoblotting using antibodies

recognising proteins MTCO1 in mitochondria, Na+/K+-ATPase in PM, IP3R3 in MAM, and

KDEL, predominantly in non-MAM ER.

119

Figure 2. Western blot analysis of subcellular fractions obtained from mouse brain. The

fractions mitochondria (Mito), endoplasmic reticulum (ER), mitochondrial associated

membranes (MAM) and plasma membrane (PM) were probed with antibodies against a)

BACE1 b) SORL1 and c) LRP/LR. d) Western immunoblotting using antibodies recognising

proteins MTCO1 mitochondria with detection in PM fraction indicating cross-contamination,

Na+/K+-ATPase in PM with detection in mitochondria fraction indicating cross-contamination,

IP3R3 in MAM, and KDEL in non-MAM ER.

120

References

1. Vance, J.E., PHOSPHOLIPID-SYNTHESIS IN A MEMBRANE-FRACTION ASSOCIATED WITH MITOCHONDRIA. Journal of Biological Chemistry, 1990. 265(13): p. 7248-7256.

2. Area-Gomez, E., et al., Upregulated function of mitochondria-associated ER membranes in Alzheimer disease. The EMBO journal, 2012. 31(21): p. 4106-23.

3. Hayashi, T. and T.P. Su, Sigma-1 receptors (sigma(1) binding sites) form raft-like microdomains and target lipid droplets on the endoplasmic reticulum: roles in endoplasmic reticulum lipid compartmentalization and export. The Journal of pharmacology and experimental therapeutics, 2003. 306(2): p. 718-25.

4. Hayashi, T. and T.P. Su, Intracellular dynamics of sigma-1 receptors (sigma(1) binding sites) in NG108-15 cells. Journal of Pharmacology and Experimental Therapeutics, 2003. 306(2): p. 726-733.

5. Simmen, T., et al., PACS-2 controls endoplasmic reticulum-mitochondria communication and Bid-mediated apoptosis. The EMBO journal, 2005. 24(4): p. 717-29.

6. de Brito, O.M. and L. Scorrano, Mitofusin 2 tethers endoplasmic reticulum to mitochondria. Nature, 2008. 456(7222): p. 605-10.

7. Rusinol, A.E., et al., A unique mitochondria-associated membrane fraction from rat liver has a high capacity for lipid synthesis and contains pre-Golgi secretory proteins including nascent lipoproteins. The Journal of biological chemistry, 1994. 269(44): p. 27494-502.

8. Chang, T.-Y., et al., Acyl-coenzyme A:cholesterol acyltransferases. American Journal of Physiology - Endocrinology and Metabolism, 2009. 297(1): p. E1-E9.

9. Stone, S.J. and J.E. Vance, Phosphatidylserine Synthase-1 and -2 Are Localized to Mitochondria-associated Membranes. Journal of Biological Chemistry, 2000. 275(44): p. 34534-34540.

10. Zborowski, J., A. Dygas, and L. Wojtczak, Phosphatidylserine decarboxylase is located on the external side of the inner mitochondrial membrane. FEBS letters, 1983. 157(1): p. 179-82.

11. Hayashi, T. and T.P. Su, Sigma-1 receptor chaperones at the ER-mitochondrion interface regulate Ca(2+) signaling and cell survival. Cell, 2007. 131(3): p. 596-610.

12. Mendes, C.C., et al., The type III inositol 1,4,5-trisphosphate receptor preferentially transmits apoptotic Ca2+ signals into mitochondria. The Journal of biological chemistry, 2005. 280(49): p. 40892-900.

13. Rizzuto, R., et al., Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses. Science, 1998. 280(5370): p. 1763-6.

14. Hamasaki, M., et al., Autophagosomes form at ER-mitochondria contact sites. Nature, 2013. 495(7441): p. 389-93.

15. Hailey, D.W., et al., Mitochondria Supply Membranes for Autophagosome Biogenesis during Starvation. Cell, 2010. 141(4): p. 656-667.

16. Simmen, T., et al., Oxidative protein folding in the endoplasmic reticulum: Tight links to the mitochondria-associated membrane (MAM). Biochimica Et Biophysica Acta-Biomembranes, 2010. 1798(8): p. 1465-1473.

17. Area-Gomez, E., et al., Presenilins are enriched in endoplasmic reticulum membranes associated with mitochondria. The American journal of pathology, 2009. 175(5): p. 1810-6.

18. Guardia-Laguarta, C., et al., alpha-Synuclein Is Localized to Mitochondria-Associated ER Membranes. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2014. 34(1): p. 249-59.

19. Newman, M., et al., Differential, dominant activation and inhibition of Notch signalling and APP cleavage by truncations of PSEN1 in human disease. Human molecular genetics, 2013.

121

20. Vassar, R., et al., Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science, 1999. 286(5440): p. 735-41.

21. Hussain, I., et al., ASP1 (BACE2) cleaves the amyloid precursor protein at the beta-secretase site. Molecular and cellular neurosciences, 2000. 16(5): p. 609-19.

22. Vassar, R., Beta-secretase (BACE) as a drug target for Alzheimer's disease. Adv Drug Deliv Rev, 2002. 54(12): p. 1589-602.

23. Zacchetti, D., et al., BACE1 expression and activity: relevance in Alzheimer's disease. Neuro-degenerative diseases, 2007. 4(2-3): p. 117-26.

24. Thinakaran, G. and E.H. Koo, Amyloid precursor protein trafficking, processing, and function. The Journal of biological chemistry, 2008. 283(44): p. 29615-9.

25. Das, U., et al., Activity-Induced Convergence of APP and BACE-1 in Acidic Microdomains via an Endocytosis-Dependent Pathway. Neuron, 2013. 79(3): p. 447-60.

26. Vassar, R., BACE1: the beta-secretase enzyme in Alzheimer's disease. Journal of molecular neuroscience : MN, 2004. 23(1-2): p. 105-14.

27. Pasternak, S.H., Presenilin-1, Nicastrin, Amyloid Precursor Protein, and -Secretase Activity Are Co-localized in the Lysosomal Membrane. Journal of Biological Chemistry, 2003. 278(29): p. 26687-26694.

28. Vassar, R., et al., The beta-Secretase Enzyme BACE in Health and Alzheimer's Disease: Regulation, Cell Biology, Function, and Therapeutic Potential. Journal of Neuroscience, 2009. 29(41): p. 12787-12794.

29. Andersen, O.M., et al., Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(38): p. 13461-13466.

30. Reitz, C., et al., Meta-analysis of the association between variants in SORL1 and Alzheimer disease. Archives of neurology, 2011. 68(1): p. 99-106.

31. Jacobsen, L., et al., Activation and functional characterization of the mosaic receptor SorLA/LR11. The Journal of biological chemistry, 2001. 276(25): p. 22788-96.

32. Rushworth, J.V., et al., Prion protein-mediated toxicity of amyloid-beta oligomers requires lipid rafts and the transmembrane LRP1. The Journal of biological chemistry, 2013. 288(13): p. 8935-51.

33. Da Costa Dias, B., et al., The 37kDa/67kDa Laminin Receptor acts as a receptor for Aβ42 internalization. Sci. Rep., 2014. 4.

34. Nikles, D., et al., Subcellular localization of prion proteins and the 37 kDa/67 kDa laminin receptor fused to fluorescent proteins. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 2008. 1782(5): p. 335-340.

122

CHAPTER 6

Research Paper 5

Latrepirdine (Dimebon) induces autophagy in

zebrafish larvae

123

124

125

Latrepirdine (Dimebon) induces autophagy in zebrafish larvae

Swamynathan Ganesan 1*Ŧ, Anne Lim 1 Ŧ, Seyyed Hani Moussavi Nik 1, Morgan

Newman 1, Prashant Bharadwaj 2, Giuseppe Verdile 2, Ralph Martins 2, 3, 4, Michael

Lardelli 1.

1. Alzheimer’s Disease Genetics Laboratory, Department of Genetics and Evolution,

School of Biological Sciences, The University of Adelaide, South Australia, Australia.

2. Centre of Excellence for Alzheimer's Disease Research and Care, School of Medical

Sciences, Edith Cowan University, Joondalup, WA, Australia.

3. School of Psychiatry and Clinical Neuroscience, University of Western Australia,

Crawley, WA, Australia.

4. The Sir James McCusker Alzheimer's Disease Research Unit, Hollywood Private

Hospital, Nedlands, WA, Australia.

SG Ŧ and AL Ŧ contributed equally to this work.

126

Abstract

Alzheimer’s disease (AD) is the most common form of dementia, and remains an

incurable disease. One of the neuropathological hallmarks AD is the presence of β-

amyloid plaques in the brain. Drug treatments for AD have thus far been unable to

significantly reverse disease progression; therefore the search for new treatments is

ongoing. Latrepirdine is one of several drugs that have been tested to combat AD.

Latrepirdine has been shown to induce autophagy and reduce β-amyloid aggregation. It

has been tested effectively on several animal models; however the effects of

Latrepirdine have not been tested on zebrafish embryos. Here we used zebrafish

embryos to test whether Latrepirdine can induce autophagy. We tested Latrepirdine at a

series of concentrations and found that a 5 µM concentration of Latrepirdine can induce

autophagy in zebrafish larvae. This is evident from a LC3 immunoblot assay, where we

observed an increase in the LC3II/LC3I ratio in the presence of a lysosomal inhibitor,

chloroquine. Our TEM analysis supported the finding that Latrepirdine induces

autophagy in zebrafish larvae at 72 hpf.

Introduction

Living organisms are under constant external stress by environmental factor. They

respond to these external factors by altering their cellular structure through various

molecular pathways. One such pathway found in most living organisms is autophagy

[1]. In this paper we discuss about macroautophagy (referred to as autophagy), which

primarily involves the degradation of proteins and organelles through the formation of

autophagosomes. The other forms of autophagy include microautophagy, which

127

involves the direct degradation of cytoplasmic cargo by lysosomes and chaperone-

mediated autophagy which involves the binding of chaperones (HSC 70) to specific

proteins and enabling their lysosomal degradation [2, 3]. Autophagy degrades proteins,

protein aggregates and whole organelles to maintain cellular homeostasis [4]. The

proteins and organelles degraded by autophagy provide substrates required for cellular

metabolism [5]. Unlike apoptosis, autophagy promotes cell survival mechanism

allowing cells to recover from damage and restore normal functioning [6]. Although

initially thought to function only in clearing harmful peptides from the cell, current

research has shown that it may have other significant functions in cellular metabolism

[7]. Studies have shown that it may play a significant role during ageing, development

and in other process which require constant protein turnover.

Recently autophagy has been implicated in cancer, neurodegenerative disorders and

other metabolic disorders [8]. Most of these disorders are multifactorial and involve

many of other factors for their progression. Therefore it becomes important to

investigate whether autophagy is protective or contributes to the progression of a

disease. Alzheimer’s disease is one of the few disorders investigated in detail for the

role of autophagy. Several studies in the past have highlighted that autophagy has a

crucial role in the progression of Alzheimer’s disease (AD), in which it may have

contradictory actions [9-11]. Some studies have shown that autophagy is impaired as a

result of AD progression [12]. Other studies have shown that autophagy may be

modulated to enhance β-amyloid clearance which is a pathological feature observed in

AD and that this may slow down disease progression [13].

128

Several drugs have been tested for their effectiveness in reducing amyloid plaque

formation [14]. Latrepirdine is one drug that has undergone clinical trials of relevance

to AD pathogenesis. Studies have shown that latrepirdine possesses neuroprotective

properties and improves cognition and behaviour in animal models [15, 16]. It has also

been shown to be involved in mitochondrial function, calcium homeostasis and other

pathways related to AD pathogenesis [17]. A study on cultured cells by Lermontova et

al (2001) has shown that Latrepirdine inhibits the neurotoxic effects of β-amyloid and

protects cells from cytotoxicity [18]. Although studies on animal models have indicated

that Latrepirdine might be useful in the treatment of cognitive disorders, the results of

clinical trials have been contradictory. An initial pilot trial and early phase II trials of

the drug showed promise, but a large scale phase III clinical trial was unsuccessful

(reviewed in [19]). The mechanism of action of Latrepirdine also remains to be fully

understood.

Zebrafish is widely used as an animal model to investigate various biochemical

pathways that contribute to AD pathogenesis [20]. Zebrafish serves as an excellent tool

to either over-express genes or to block gene expression by using morpholino antisense

oligonucleotides [21, 22]. Zebrafish embryos are optically transparent and chemicals

can be directly added to their aqueous support medium making it easy to analyse their

effects on various biochemical pathways [23, 24]. Previously we have shown that

autophagy can be analysed in zebrafish embryos by using various chemicals that induce

and inhibit this function. We have also validated that autophagy can be monitored

effectively in zebrafish embryos using the LC3 immunoblot assay [25].

129

In the work described in this paper, we use 3-day old zebrafish larvae to test whether

Latrepirdine (Dimebon) induces autophagy. Although Latrepirdine has been shown to

induce autophagy in cell lines, we report here the inductive effect of latrepirdine on

autophagy in zebrafish. We show that autophagy induced by Latrepirdine can be

monitored more sensitively in zebrafish embryos by direct observation of

autophagosomes by transmission electron microscopy (TEM). Our findings support the

use of zebrafish as a system in which to investigate the effects of Latrepirdine in vivo.

Materials and Methods

Ethics

This work was conducted under the auspices of The Animal Ethics Committee of The

University of Adelaide and in accordance with EC Directive 86/609/EEC for animal

experiments and the Uniform Requirements for Manuscripts Submitted to Biomedical

Journals.

Zebrafish husbandry and experimental procedures

Wild-type zebrafish were maintained in a recirculated water system with a 14 hour

light/10 hour dark cycle. Fertilized embryos were placed in petri plates containing

embryo medium (E3) and grown at 28 ˚C in a humid incubator.

130

Drug Treatment

Latrepirdine (Dimebon) was added to 66 hpf larvae for 6 h at the concentrations

specified in the experimental descriptions (1 µM to 200 µM). Chloroquine, a lysosomal

inhibitor was added to a final concentration of 50 µM at 56 hpf followed by incubation

for a further 16 h [23, 26]. At 72 hpf, all larvae were deyolked. The deyolked samples

were then lysed with sample lysis buffer and immunoblotted.

Western immunoblot analyses

Deyolked larvae were placed in sample buffer (2% sodium dodecyl sulfate (SDS), 5%

β-mercaptoethanol, 25% v/v glycerol, 0.0625 M Tris–HCl (pH 6.8), and bromphenol

blue), and then heated immediately at 100 ºC for 5 min. Samples were run on 15%

SDS-PAGE gels and proteins were transferred to PVDF membranes using a wet

electrotransfer system. The membranes were blocked with 5% Western Blocking

Solution in TBST and then probed with anti-LC3 antibody and anti-β-tubulin antibody

(Antibody E7, Developmental Studies Hybridoma Bank, The University of Iowa, IA,

USA) which was used as a loading control. The blots were visualized using luminol

reagents (Amresco, Ohio, USA) and the ChemiDoc™ MP imaging system (Bio-Rad,

Hercules, CA, USA).

Transmission electron microscopy

To observe the number of autophagosomes in zebrafish cells, treated and untreated

larvae were grown to the 72 hpf stage. The larvae were fixed in 4% paraformaldehyde,

131

1.25% glutaraldehyde in PBS with 4% sucrose, pH 7.2 overnight at 4 ˚C, and then

rinsed in washing buffer (PBS with 4% sucrose) twice for ten minutes. Larvae were

post-fixed in 2% osmium tetroxide for 45 minutes on a rotator, followed by dehydration

through an ethanol series of 70, 90, 95 and 100%. The ethanol was replaced by

propylene oxide for 15 minutes followed by a mixture of propylene oxide and resin at a

1:1 ratio for 45 minutes. This was followed by 100% resin infiltration overnight. The

embryos were then embedded in fresh resin and the resin polymerized at 70 ˚C for 24

hours. The embedded larvae were sectioned coronally in a rostral to caudal direction

through the head of the embryo using an ultramicrotome. The depth at which sections

were obtained was approximately 100-130 μm from the rostral end of the head, roughly

one-fifth into the eye to obtain an ultrathin 85 nm section of the forebrain. The ultrathin

sections were stained with uranyl acetate and lead citrate and imaged on an Olympus-

SIS Veleta CCD camera in a FEI Tecnai G2 Spirit TEM (FEI, Oregon, USA). Images

were obtained from 10 cells per larva. Thus, the cells observed for counting of

autophagosomes were 30 per treatment.

Statistical analysis

Each experiment was carried out with 3 biological replicates (n=3). Data were analysed

using GraphPad Version 6.0 (GraphPad Prism, La Jolla, CA). For the LC3II/LC3I

assays, since single clutches of larvae were used to generate the various treatments of

any biological replicate, differences among the groups were evaluated using a two-

tailed paired t-test. For the TEM analysis, the treatments were conducted on one clutch

of larvae and cells from three separate larvae from each treatment were analysed. P-

values were calculated for all the comparisons and p<0.05 was considered significant.

132

Results

Determining the concentration of Latrepirdine that can induce autophagy in

zebrafish embryos by an LC3 immunoblot assay

The most commonly used assay to measure autophagy is monitoring of the conversion

of LC3, an autophagy marker. LC3 is a microtubule associated protein that is present in

cytosol as LC3I and LC3II. Upon induction of autophagy, the cytosol LC3I undergoes

lipidation with phosphotidyl ethanolamine (PE) and converts to LC3II which then binds

to autophagosomes. Therefore determining the ratio of LC3II to LC3I reveals changes

in autophagic activity.

Latrepirdine has been shown to induce autophagy in cell lines and in mice but no study

has shown whether this effect can be generalised to other in vivo models. Since we are

interested in the use of zebrafish to study Alzheimer’s disease, we carried out an initial

screening of Latrepirdine at different concentrations to find out an optimum

concentration that might induce autophagy in zebrafish larvae. We performed a

standard LC3 Immunoblot assay on 72 hpf zebrafish embryos treated with a range of

concentrations of Latrepirdine. In our initial screening we used Latrepirdine

concentrations ranging from 50µM to 200µM. We then calculated the LC3II/LC3I ratio

and did not find any significant difference in the ratio between control and Latrepirdine-

treated samples (Figure 1B). Previous studies have demonstrated that a 10 nM

concentration of Latrepirdine can induce autophagy in SH-SY5Y cells [27]. A very

high concentration of Latrepirdine could be a reason for not finding any significant

change and we decided to do a screen with lower concentrations of Latrepirdine.

133

In our second screen, we used Latrepirdine at concentrations in the range of 1µM to

50µM and performed the LC3 immunoblot assay. We found the most significant

increase in the LC3II/LC3I ratio was observed between samples treated with 5 µM

Latrepirdine and untreated control samples at 72 hpf. We saw that the autophagic

induction increases gradually with increasing Latrepirdine concentration and reaches a

maximum increase at 5 µM of Latrepirdine when compared to untreated wild-type

larvae (Figure 2B).

Currently, a standard requirement to validate observations of changes in LC3 isoform

ratios is to observe the effects on LC3II of lysosomal inhibition. (If a treatment has

increased autophagic flux then lysosomal inhibition should result in relative LC3II

accumulation). Therefore we performed the LC3 immunoblot assay in the presence of

chloroquine, a lysosomal inhibitor to validate our findings further.

Induction of autophagy by Latrepirdine is confirmed by an increase in the LC3II/I

ratio in the presence of chloroquine

The LC3 immunoblot assay was performed using the 5 µM concentration of

Latrepirdine that had shown an increase in the LC3II/LC3I ratio. No significant

morphological changes were observed when larvae were treated with 5 µM Latrepirdine

and chloroquine either in combination or alone suggesting that these treatments are not

toxic (Figure 3). Larvae treated with 5 µM Latrepirdine and chloroquine showed a

significant increase in the LC3II/LC3I ratio when compared to wild-type larvae treated

134

with chloroquine alone. This supported our finding that 5 µM Latrepirdine induces

autophagy in zebrafish larvae (Figure 4B).

Determining induction of autophagy by counting the number of autophagic

vacuoles using transmission electron microscopy (TEM)

Autophagosomes were first discovered using electron microscopy [28]. Transmission

electron microscopy (TEM) is a widely used technique to assess numbers of autophagic

vacuoles in cells. Electron microscopy produces high resolution images of

autophagosomes which give a visual validation of the presence of autophagic vesicles

assessed in other assays such as using the LC3 immunoblot assay. In electron

microscopy autophagosomes can be identified as membrane bound vesicles that contain

cytosolic materials such as organelles [29]. Our LC3 immunoblot assay showed that

autophagy can be induced by Latrepirdine in zebrafish larvae. In order to validate

further our LC3 immunoblot assay findings, we used TEM to quantitate the number of

autophagic vacuoles present in the brains of treated and untreated zebrafish larvae. To

do this we treated larvae using the same conditions used for the immunoblot assay and

then processed the larvae for TEM. All embryos were sectioned in the forebrain region,

approximately 100-130 μm from the rostral extremity of the head. We chose cells from

brain sections because LC3 levels are known to be high in the brain and Latrepirdine is

proposed as a treatment for the neurodegenerative disease AD. We used the presence of

the mandibular cartilage and ethmoid plate of the larva as a visual guide to confirm that

we were observing cells from similar regions of each larva (Figure 5). Cells from the

centre of the ventral area of the forebrain were chosen since these cells appeared fairly

uniform in morphology and may represent a more homogenous cell type (Figure 5).

135

We analysed the number of autophagosomes present in ten cells per zebrafish larva and

repeated this in triplicate, using three biological replicates. Cells of untreated larvae

rarely contained autophagosomes, but treatment of larvae with only chloroquine

increased the number of autophagosomes in cells by more than three-fold (Figure 6).

This appears to indicate that under normal conditions, autophagosomes are degraded

rapidly after induction in larvae and chloroquine treatment blocks their degradation by

inhibiting the activity of lysosomal enzymes. We found that treatment with 5 µM

Latrepirdine significantly increased the number of autophagosomes present in cells

when compared with untreated larvae (Figure 7). For a more accurate assessment of

autophagic flux, we compared larvae treated with 5 µM Latrepirdine in the presence of

chloroquine to larvae treated with chloroquine only as was done in our immunoblot

analyses. We found that Latrepirdine significantly increased the number of autophagic

vacuoles present in cells (Figure 7). Our TEM results are in agreement with the LC3

immunoblot assay and indicate that 5 µM Latrepirdine causes induction of autophagy in

the brains of zebrafish larvae at 72 h.

Discussion

AD is a complex multifactorial disorder characterised by the accumulation of β-amyloid

in plaques in the brain [30-32]. β-amyloid accumulation is one of the important

pathological features observed during the progression of the disease, therefore much

research has focussed on the neurotoxicity of β-amyloid aggregation. Research has also

identified several other factors influencing AD progression [33]. One such factor is

autophagy. It is thought that failure of β-amyloid clearance due to a block in the

autophagy pathway might play a critical role in AD progression [34, 35]. This has led to

136

studies of several compounds that can possibly induce autophagy and might potentially

alleviate AD symptoms [36].

In recent times, research has focussed on identifying potential drugs that might alleviate

AD symptoms by reducing β-amyloid aggregation through stimulation of autophagy.

Latrepirdine is one such drug that has been tested for amyloid clearance on yeast and

mouse models [37, 38]. Initially studied as an anti-histamine drug, subsequent research

has tested whether Latrepirdine has any other neuroprotective properties [39]. Studies

have shown promising results regarding the effectiveness of Latrepirdine for reducing

symptoms in several neurodegenerative disorders. While some research carried out on

Latrepirdine has suggested that it acts as a neuroprotector and enhances improvement in

cognition and behaviour [15, 40, 41], other studies have shown that it can also reduce

the aggregation of β-amyloid [42]. Moreover it has also been shown to interact with

several other pathways involved in neurodegenerative disorders [43]. Although

Latrepirdine has been shown to have neuroprotective properties, the pathways and

mechanisms it interacts with to achieve this are still unknown. Therefore further

research is required to characterize this chemical and to understand the various critical

pathways influenced by it.

In our study, we examined whether Latrepirdine can induce autophagy in zebrafish

larvae. Zebrafish is an excellent tool for dissection of molecular pathways as embryos

and, to some extent, larvae can be manipulated to over-express or inhibit gene activity

[44]. Also chemicals can be directly added to the fish’s support medium making it

137

simple to observe effects at the molecular level and in animal behaviour. The ease of

breeding large numbers of zebrafish embryos in a short time span and its cost

effectiveness compared to other animal models such as mice make zebrafish a popular

model for testing drugs such as Latrepirdine. Using animal models with a vertebrate

system is more advantageous over organisms such as yeast as more complex

interactions within a particular type of cell, tissue or organ can be observed. Therefore

the zebrafish has the benefits of higher organism animal models while also being easy

to use much like a lower organism. Furthermore, 70% of human genes have at least one

clear zebrafish orthologue, including genes involved in autophagy [45-48]. Therefore it

would seem that zebrafish are the optimum animal model to use for further studies on

Latrerpirdine and other autophagy modifying drugs.

One aim of our study was to test whether treatment with Latrepirdine has any

morphological effects on early zebrafish development since this may indicate drug

toxicity. Latrepirdine treatment of developing zebrafish embryos did not result in any

gross morphological changes. Another aim was to find an optimum concentration of

Latrepirdine for induction of autophagy in zebrafish larvae. Through our concentration

screening we identified that Latrepirdine at 5 µM concentration induces autophagy in

zebrafish larvae by 72 h. Further testing of 5 µM Latrepirdine in the presence of the

lysosomal inhibitor, chloroquine, confirmed the induction of autophagy by this drug. To

support the findings of our LC3 immunoblot assay we performed a TEM analysis. The

TEM analysis allowed actual quantification of autophagosomes and supported the

findings of the LC3 assays. Future analysis will focus on identifying the underlying

molecular machinery of autophagy that is affected by Latrepirdine.

138

In conclusion we have shown that exposure of developing zebrafish larvae to 5 µM

Latrepirdine can induce autophagy in zebrafish by 72 hpf. Our observations are

consistent with other published studies showing that Latrepirdine can induce autophagy

in a number of other animal models and cell lines. The activity of Latrepirdine in

zebrafish opens up several avenues to understanding and testing the properties of this

drug with respect to AD progression and it is hoped that further understanding of the

mechanism of action of Latrepirdine will be elucidated. This also supports the use of

zebrafish in testing of other drugs that may have potential in treatment of AD.

139

Figures

Figure 1 – (A) LC3 Immunoblot assay to analyse the effect on autophagy of

concentrations of 50 µM-200 µM of Latrepirdine. Zebrafish larvae were treated with

the Latrepirdine concentrations shown from 66 hpf before deyolking at 72 h, lysis and

SDS-PAGE followed by detection of LC3II and LC3I using an anti-LC3 antibody.

Anti-tubulin antibody was used as a loading control. (B) The LC3II/LC3I ratio was

calculated using Image Lab software. Each analysis was carried out in triplicate. Error

bars represent standard error of means (n=3).

140

Figure 2 – (A) LC3 Immunoblot assay to analyse the effect on autophagy of

concentrations of 1 µM-50 µM of Latrepirdine. Larvae were treated with the

Latrepirdine concentrations shown from 66 hpf. Larvae were then deyolked at 72 h,

lysed and subjected to SDS-PAGE before LC3II and LC3I were detected by

immunoblotting using an anti-LC3 antibody. Anti-tubulin antibody was used as a

loading control. (B) The LC3II/LC3I ratio was calculated using Image Lab software.

Each analysis was carried out in triplicate. Error bars represent standard errors of

means.

141

Figure 3 – Dissection microscope images of zebrafish larvae at 72hpf showing the

absence of morphological changes under 5 µM Latrepirdine and chloroquine treatment.

Representative image of respective treatments are shown. CQ, Chloroquine;

142

Figure 4 – LC3 Immunoblot assay carried in the presence of chloroquine, a lysosomal

inhibitor. Larvae treated with 5 µM Latrepirdine and chloroquine were deyolked at 72

h, lysed and subjected to SDS-PAGE before LC3II and LC3I were detected by

immunoblotting using an anti-LC3 antibody. Anti-tubulin antibody was used as a

loading control. (B) The LC3II/LC3I ratio was calculated using Image Lab software.

Each analysis was carried out in triplicate. Error bars represent standard errors of

means. CQ, Chloroquine;

143

Figure 5 – TEM image showing a transverse section through the head of a zebrafish

larva at 72 h. The ventral forebrain in which cells were imaged in our TEM studies is

indicated by asterisks (*) within the boxed region. The ethmoid plate (E) and

mandibular cartilage (M) were used as visual references to locate a similar area of the

brain for imaging in the various treatment samples and replicates. L, lens; IPL, inner

plexiform layer; OPL, outer plexiform layer; GCL, ganglion cell layer;

144

Figure 6 - TEM of autophagic vacuoles present in a cell. Electron microscopy images

from cells from the ventral forebrain of larvae treated with (a) control (b) 5 μM

Latrepirdine (c) control with chloroquine (d) 5 μM Latrepirdine with chloroquine. Each

image shows the presence or absence of autophagic vacuoles in a cell from the brain of

a 72 hpf larva. Arrows indicate autophagic vacuoles. The control cell, panel (a) does not

contain any autophagic vacuoles. Panel (d) is a composite of two images of a single cell

as autophagic vacuoles were located in different regions within the cell.

145

Figure 7 - The quantification of autophagosomes using TEM indicates that Latrepirdine

increases the number of autophagosomes present in cells in the ventral forebrain of

zebrafish larvae. Autophagosomes were counted in ten cells per embryo and the

experiment was done using three separate larvae (n=3). Error bars represent standard

deviation, statistical significance values are from a two-tailed unpaired student’s t-test

assuming unequal variance.

146

References

1. Rikiishi, H., Novel Insights into the Interplay between Apoptosis and Autophagy. Int J Cell Biol, 2012. 2012: p. 317645.

2. Boya, P., F. Reggiori, and P. Codogno, Emerging regulation and functions of autophagy. Nat Cell Biol, 2013. 15(7): p. 713-20.

3. Reggiori, F., et al., Selective types of autophagy. Int J Cell Biol, 2012. 2012: p. 156272. 4. Mizushima, N. and B. Levine, Autophagy in mammalian development and

differentiation. Nat Cell Biol, 2010. 12(9): p. 823-30. 5. Parzych, K.R. and D.J. Klionsky, An overview of autophagy: morphology, mechanism,

and regulation. Antioxid Redox Signal, 2014. 20(3): p. 460-73. 6. Altman, J.K., et al., Autophagy is a survival mechanism of acute myelogenous leukemia

precursors during dual mTORC2/mTORC1 targeting. Clin Cancer Res, 2014. 20(9): p. 2400-9.

7. Kim, K.H. and M.S. Lee, Autophagy--a key player in cellular and body metabolism. Nat Rev Endocrinol, 2014. 10(6): p. 322-37.

8. Shintani, T. and D.J. Klionsky, Autophagy in health and disease: a double-edged sword. Science, 2004. 306(5698): p. 990-5.

9. Tung, Y.T., et al., Autophagy: a double-edged sword in Alzheimer's disease. J Biosci, 2012. 37(1): p. 157-65.

10. Boland, B., et al., Autophagy induction and autophagosome clearance in neurons: relationship to autophagic pathology in Alzheimer's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2008. 28(27): p. 6926-37.

11. Yang, D.S., J.H. Lee, and R.A. Nixon, Monitoring autophagy in Alzheimer's disease and related neurodegenerative diseases. Methods in enzymology, 2009. 453: p. 111-44.

12. Li, L., X. Zhang, and W. Le, Autophagy dysfunction in Alzheimer's disease. Neurodegener Dis, 2010. 7(4): p. 265-71.

13. Zhu, X.C., et al., Autophagy modulation for Alzheimer's disease therapy. Mol Neurobiol, 2013. 48(3): p. 702-14.

14. Nakamura, Y., [New anti-AD drugs--their possibilities and issues]. Seishin Shinkeigaku Zasshi, 2012. 114(3): p. 255-61.

15. Bachurin, S., et al., Antihistamine agent Dimebon as a novel neuroprotector and a cognition enhancer. Ann N Y Acad Sci, 2001. 939: p. 425-35.

16. Lermontova, N.N., et al., Dimebon improves learning in animals with experimental Alzheimer's disease. Bull Exp Biol Med, 2000. 129(6): p. 544-6.

17. Wu, J., Q. Li, and I. Bezprozvanny, Evaluation of Dimebon in cellular model of Huntington's disease. Mol Neurodegener, 2008. 3: p. 15.

18. Lermontova, N.N., et al., Dimebon and tacrine inhibit neurotoxic action of beta-amyloid in culture and block L-type Ca(2+) channels. Bull Exp Biol Med, 2001. 132(5): p. 1079-83.

19. Bezprozvanny, I., The rise and fall of Dimebon. Drug news & perspectives, 2010. 23(8): p. 518-523.

20. Newman, M., et al., Zebrafish as a tool in Alzheimer's disease research. Biochim Biophys Acta, 2011. 1812(3): p. 346-52.

21. Nornes, S., et al., Interference with splicing of Presenilin transcripts has potent dominant negative effects on Presenilin activity. Hum Mol Genet, 2008. 17(3): p. 402-12.

22. Nornes, S., et al., Independent and cooperative action of Psen2 with Psen1 in zebrafish embryos. Exp Cell Res, 2009. 315(16): p. 2791-801.

147

23. Ganesan, S., et al., Identification and expression analysis of the zebrafish orthologues of the mammalian MAP1LC3 gene family. Exp Cell Res, 2014.

24. Moussavi Nik, S.H., et al., The comparison of methods for measuring oxidative stress in zebrafish brains. Zebrafish, 2014. 11(3): p. 248-54.

25. Ganesan, S., et al., Identification and expression analysis of the zebrafish orthologues of the mammalian MAP1LC3 gene family. Exp Cell Res, 2014. 328(1): p. 228-37.

26. Cui, J., et al., Generation of transgenic zebrafish with liver-specific expression of EGFP-Lc3: a new in vivo model for investigation of liver autophagy. Biochem Biophys Res Commun, 2012. 422(2): p. 268-73.

27. Steele, J.W., et al., Latrepirdine stimulates autophagy and reduces accumulation of alpha-synuclein in cells and in mouse brain. Mol Psychiatry, 2013. 18(8): p. 882-8.

28. Novikoff, A.B., The proximal tubule cell in experimental hydronephrosis. J Biophys Biochem Cytol, 1959. 6(1): p. 136-8.

29. Eskelinen, E.L., Maturation of autophagic vacuoles in Mammalian cells. Autophagy, 2005. 1(1): p. 1-10.

30. Capetillo-Zarate, E., et al., Intraneuronal Abeta accumulation, amyloid plaques, and synapse pathology in Alzheimer's disease. Neurodegener Dis, 2012. 10(1-4): p. 56-9.

31. Yilmazer-Hanke, D.M., Alzheimer's disease. The density of amygdalar neuritic plaques is associated with the severity of neurofibrillary pathology and the degree of beta-amyloid protein deposition in the cerebral cortex. Acta Anat (Basel), 1998. 162(1): p. 46-55.

32. Mann, S.S. and J.A. Hammarback, Molecular characterization of light chain 3. A microtubule binding subunit of MAP1A and MAP1B. J Biol Chem, 1994. 269(15): p. 11492-7.

33. Munoz, D.G. and H. Feldman, Causes of Alzheimer's disease. CMAJ, 2000. 162(1): p. 65-72.

34. Nixon, R.A. and D.S. Yang, Autophagy failure in Alzheimer's disease--locating the primary defect. Neurobiol Dis, 2011. 43(1): p. 38-45.

35. Wolfe, D.M., et al., Autophagy failure in Alzheimer's disease and the role of defective lysosomal acidification. Eur J Neurosci, 2013. 37(12): p. 1949-61.

36. Yang, D.S., et al., Therapeutic effects of remediating autophagy failure in a mouse model of Alzheimer disease by enhancing lysosomal proteolysis. Autophagy, 2011. 7(7): p. 788-9.

37. Bharadwaj, P.R., et al., Latrepirdine (dimebon) enhances autophagy and reduces intracellular GFP-Abeta42 levels in yeast. J Alzheimers Dis, 2012. 32(4): p. 949-67.

38. Steele, J.W. and S. Gandy, Latrepirdine (Dimebon(R)), a potential Alzheimer therapeutic, regulates autophagy and neuropathology in an Alzheimer mouse model. Autophagy, 2013. 9(4): p. 617-8.

39. Okun, I., et al., From anti-allergic to anti-Alzheimer's: Molecular pharmacology of Dimebon. Curr Alzheimer Res, 2010. 7(2): p. 97-112.

40. Doody, R.S., et al., Effect of dimebon on cognition, activities of daily living, behaviour, and global function in patients with mild-to-moderate Alzheimer's disease: a randomised, double-blind, placebo-controlled study. Lancet, 2008. 372(9634): p. 207-15.

41. Egea, J., et al., Neuroprotective effect of dimebon against ischemic neuronal damage. Neuroscience, 2014. 267: p. 11-21.

42. Ustyugov, A.A., et al., Dimebon reduces the levels of aggregated amyloidogenic protein forms in detergent-insoluble fractions in vivo. Bull Exp Biol Med, 2012. 152(6): p. 731-3.

148

43. Zhang, S., et al., Dimebon (latrepirdine) enhances mitochondrial function and protects neuronal cells from death. J Alzheimers Dis, 2010. 21(2): p. 389-402.

44. Newman, M., E. Ebrahimie, and M. Lardelli, Using the zebrafish model for Alzheimer's disease research. Front Genet, 2014. 5: p. 189.

45. Howe, K., et al., The zebrafish reference genome sequence and its relationship to the human genome. Nature, 2013. 496(7446): p. 498-503.

46. Makky, K., J. Tekiela, and A.N. Mayer, Target of rapamycin (TOR) signaling controls epithelial morphogenesis in the vertebrate intestine. Developmental biology, 2007. 303(2): p. 501-13.

47. Tsukamoto, S., et al., Autophagy is essential for preimplantation development of mouse embryos. Science, 2008. 321(5885): p. 117-20.

48. He, C., et al., Assaying autophagic activity in transgenic GFP-Lc3 and GFP-Gabarap zebrafish embryos. Autophagy, 2009. 5(4): p. 520-6.

149

Supplementary Data

Table 1 – Densitometric analysis of western blots using zebrafish larvae (control

larvae and larvae treated with concentrations of 50 µM - 200 µM of Latrepirdine

(Figure 1B). The intensity values of the LC3I and LC3II bands of each replicate

are given.

Set 1 Control 50µM 100µM 125µM 150µM LC3I 2,924,565 3,066,416 3,050,582 4,251,522 4,596,696 LC3II 8,201,865 7,402,063 8,745,948 17,500,032 13,469,248

LC3II/LC3I Ratio 2.804473486 2.413913507 2.866976859 4.116180511 2.930202041

Set 2 Control 50µM 100µM 125µM 150µM LC3I 2,798,684 1,803,822 2,384,046 2,401,584 2,161,632 LC3II 10,905,293 11,003,508 12,686,490 10,768,800 8,330,112

LC3II/LC3I Ratio 3.896578892 6.100107439 5.321411584 4.484040533 3.853621708

Set 3 Control 50µM 100µM 125µM 150µM LC3I 4,933,026 6,504,685 8,707,920 9,521,875 8,195,550 LC3II 11,437,027 14,438,655 20,533,812 25,887,840 21,719,350

LC3II/LC3I Ratio 2.318460718 2.219731624 2.358061627 2.718775451 2.650139405

Set 1 175µM 200µM LC3I 4,325,832 5,720,767 LC3II 5,133,618 14,662,662

LC3II/LC3I Ratio 1.186735407 2.563058765

Set 2 175µM 200µM LC3I 4,373,692 3,965,998 LC3II 16,447,500 12,690,068

LC3II/LC3I Ratio 3.760552869 3.199716187

Set 3 175µM 200µM LC3I 13,653,935 7,232,114 LC3II 28,555,259 19,131,712

LC3II/LC3I Ratio 2.091357473 2.645383079

150

Statistical Analysis p-value Control Vs 50µM 0.5582 Control Vs 100µM 0.3821 Control Vs 125µM 0.1102 Control Vs 150µM 0.3303 Control Vs 175µM 0.3023 Control Vs 200µM 0.5625

151

Table 2 – Densitometric analysis of western blots using zebrafish larvae (control

larvae and larvae treated with concentrations of 1 µM - 50 µM of latrepirdine

(Figure 2B). The intensity values of the LC3I and LC3II bands of each replicate

are given.

Set 1 Control 1µM 2.5µM 5µM 10µM 50µM LC3I 6,372,826 5,346,250 9,046,125 3,505,680 5,724,950 6,585,696 LC3II 4,425,341 5,253,942 9,433,113 4,535,028 6,295,100 8,033,088

LC3II/LC3I Ratio

0.694407944

0.982734066

1.042779422

1.293622921

1.099590389

1.219778137

Set 2 Control 1µM 2.5µM 5µM 10µM 50µM

LC3I 8,845,912 8,648,350 7,465,635 8,320,884 10,600,79

5 6,663,519

LC3II 13,098,47

3 15,102,15

0 15,076,48

5 19,418,88

0 20,593,36

2 11,166,40

1

LC3II/LC3I Ratio

1.48073743

1.746246394

2.019451125

2.333752039

1.942624303

1.675751356

Set 3 Control 1µM 2.5µM 5µM 10µM 50µM LC3I 2,588,572 3,867,540 3,040,477 4,953,832 3,415,808 3,487,146

LC3II 6,558,051 10,069,68

0 8,627,743 16,778,63

8 9,748,244 11,419,09

5

LC3II/LC3I Ratio

2.533462851

2.603639523

2.837628109

3.38700182

2.85386181

3.274624865

152

Statistical Analysis p-value Control Vs 1µM 0.0952 Control Vs 2.5µM 0.0313 Control Vs 5µM 0.0119 Control Vs 10µM 0.0106 Control Vs 50µM 0.0919

153

Table 3 – Densitometric analysis of western blots of zebrafish larvae (control

larvae and larvae treated with 5 µM Latrepirdine and 50 µM chloroquine (Figure

4B). The intensity values of the LC3I and LC3II bands of each replicate are given.

Set 1 Con+Chl Lat(5µM)+Chl LC3I 724,458 560,306 LC3II 2,606,352 2,379,271

LC3II/LC3I Ratio 3.59765784 4.246377872

Set 2 Con+Chl Lat(5µM)+Chl LC3I 605,324 565,600 LC3II 2,077,921 2,264,800

LC3II/LC3I Ratio 3.4327418 4.004243281

Set 3 Con+Chl Lat(5µM)+Chl LC3I 676,623 1,356,390 LC3II 1,198,717 3,056,382

LC3II/LC3I Ratio 1.77161728 2.253320948

Statistical Analysis p-value Con+Chl Vs Lat(5µM)+Chl 0.0072

154

Table 4 – Counting of autophagosomes in 10 cells from the ventral forebrain of a

72hpf zebrafish larva. 3 biological replicates were performed using TEM (Figure

7) Lat, Latrepirdine; CQ, Chloroquine;

Control

Control+CQ

Lat

Lat + CQ Sample 1

Sample 1

Sample 1

Sample 1

Cell 1 0

Cell 1 1

Cell 1 1

Cell 1 2 Cell 2 0

Cell 2 1

Cell 2 1

Cell 2 3

Cell 3 0

Cell 3 1

Cell 3 1

Cell 3 4 Cell 4 0

Cell 4 1

Cell 4 4

Cell 4 2

Cell 5 0

Cell 5 1

Cell 5 1

Cell 5 3 Cell 6 1

Cell 6 1

Cell 6 2

Cell 6 3

Cell 7 0

Cell 7 0

Cell 7 0

Cell 7 5 Cell 8 2

Cell 8 1

Cell 8 1

Cell 8 0

Cell 9 0

Cell 9 0

Cell 9 0

Cell 9 3 Cell 10 0

Cell 10 1

Cell 10 2

Cell 10

Average 3

8

13

25

Sample 2

Sample 2

Sample 2

Sample 2 Cell 1 0

Cell 1 1

Cell 1 1

Cell 1 2

Cell 2 1

Cell 2 1

Cell 2 1

Cell 2 4 Cell 3 0

Cell 3 0

Cell 3 2

Cell 3 3

Cell 4 0

Cell 4 3

Cell 4 1

Cell 4 2 Cell 5 1

Cell 5 2

Cell 5 2

Cell 5 1

Cell 6 1

Cell 6 1

Cell 6 1

Cell 6 1 Cell 7 0

Cell 7 0

Cell 7 2

Cell 7 0

Cell 8 0

Cell 8 0

Cell 8 0

Cell 8 1 Cell 9 0

Cell 9 1

Cell 9 0

Cell 9 2

Cell 10 0

Cell 10 2

Cell 10 1

Cell 10 2 Average 3

11

11

18

Sample 3

Sample 3

Sample 3

Sample 3 Cell 1 0

Cell 1 0

Cell 1 1

Cell 1 3

Cell 2 0

Cell 2 0

Cell 2 1

Cell 2 2 Cell 3 0

Cell 3 2

Cell 3 1

Cell 3 3

Cell 4 0

Cell 4 1

Cell 4 3

Cell 4 0 Cell 5 0

Cell 5 1

Cell 5 3

Cell 5 1

155

Cell 6 0

Cell 6 1

Cell 6 1

Cell 6 3 Cell 7 0

Cell 7 1

Cell 7 2

Cell 7 2

Cell 8 0

Cell 8 0

Cell 8 1

Cell 8 1 Cell 9 1

Cell 9 1

Cell 9 0

Cell 9 3

Cell 10 0

Cell 10 0

Cell 10 1

Cell 10 3 Average 1

7

14

21

Total Number of Autophagosomes

Control 7 Control + Chloroquine (50µM) 26 Latrepirdine (5µM) 38 Latrepirdine (5µM) + Chloroquine (50µM) 64

Statistical Analysis p-value Control Vs Lat 0.001 Control Vs Con+CQ 0.0176 Lat Vs Lat+CQ 0.035 Con+CQ Vs Lat+CQ 0.0102

156

CHAPTER 7

DISCUSSION

157

AD is a complex disease with an unclear underlying pathology. Much focus has remained on the

PRESENILINS, mainly PSEN1 due to its role as the catalytic core of the γ-secretase complex

producing Aβ, as well as the many mutations in PSEN1 that have been linked to FAD.

Understanding the role PSENs play in AD could provide insight to the mechanisms of both SAD

and FAD. Furthermore, focus on the MAM sub-compartment, where PSENs are enriched, may

reveal more clearly the pathology of AD. However, understanding the structure and function of

the other components of γ-secretase is equally important to fully comprehend how mutations in

PSEN1 potentially affect the complex as a whole. NCSTN is one of the four main components of

the γ-secretase complex. In zebrafish, the psen1, psen2, aph1 and psenen genes have been studied

and characterized to some extent, but information regarding zebrafish ncstn is lacking.

In this thesis we have confirmed the identity of zebrafish ncstn. The conservation of the

glutamate residue cognate to E333 in human and the DYIGS motif that are important for either

Ncstn substrate recognition or γ-secretase formation indicate that zebrafish Ncstn likely retains

the same functions as human NCSTN. This thesis includes the first detailed publication regarding

zebrafish Ncstn and hopefully has laid the groundwork to further clarify Ncstn’s function using

the zebrafish model. It would be interesting to further analyze why ncstn expression levels (high)

do not coordinate with those of psen1 and psen2 (low) in adult tissues.

To date more than 170 mutations have been identified in the PSEN1 gene

(http://www.molgen.ua.ac.be/admutations). Some mutations cause aberrant splicing or result in a

frame shift leading to a premature stop codon and transcripts encoding truncated proteins [34, 72,

158

221]. Various truncations of human PSEN1 and zebrafish Psen1 have been shown to have

differential effects on Notch signalling and Appa cleavage [219] but the mechanisms by which

these truncations exert their effects remain unclear. One possibility was that these different

effects were due to the different sizes of truncated protein localizing to different cellular

compartments.

We demonstrated that EGFP-tagged truncated Presenilin1 proteins were observed to be

uniformly distributed in the cytoplasm but not imported into the nucleus, with a lack of EGFP

fluorescence in lysosomes. However it cannot be ruled out that this may have been due to the pH

sensitivity of EGFP and protein degradation. Therefore further work to confirm the localisation

of Psen1 truncations will help to validate the results seen here. Although preliminary, our results

in this thesis indicate that the different truncations analysed appear to have similar distributions,

therefore their differential effects on γ-secretase substrates are likely not due to their different

subcellular localisations.

Another possibility is that the different sizes of truncated proteins can bind to and interfere with

wild type Psen1 function either within the γ-secretase complex or as a holoprotein. It is possible

that the structure of each truncated protein may differentially change the arrangement of the four

components of the γ-secretase complex structure within the bilayer, thus having different effects

on different substrates. It has been shown that wild type Psen1 is required for the effects of one of

the truncations to be seen [219]. Previous work has suggested that changes to membrane

thickness in which γ-secretase complex is localized could interfere in the packing and tilting of its

159

transmembrane domains and might influence the positioning of the substrate cleavage site

thereby changing the ratio of Aβ40 to Aβ42 produced [222]. Therefore it could be possible that the

truncated forms of Psen1 may also have a similar effect on substrate cleavage. Analysing the

three dimensional structure of the γ-secretase complex or holoprotein obtained from the mRNA-

injected embryos, using cryo-EM single-particle analysis (as performed in [62]) would be a valid

method to observe any differences in structure.

This thesis also gives focus to the ER subcompartment which interacts physically and

biochemically with mitochondria, the MAM. We investigated the potential of the zebrafish model

as a suitable system in which to study the effects of Psen1 and Psen2 activity on mitochondrion-

ER apposition. Our findings indicate that blocking translation of the Psen proteins using

morpholinos in single-cell stage embryos results in no significant change in the distribution of

mitochondrion-ER apposition lengths at 24 hours post fertilization. Therefore we could not

replicate the observations of increased lengths between ER and mitochondria that were seen in

mammalian cells lacking PSEN1 and PSEN2. There are a number of possible reasons for this

difference, with the most likely being an inability to reduce Psen activity sufficiently in embryos

at the age of 24 hours. The expression of mRNAs of genes encoding the components of the γ-

secretase complex are contributed to early embryos maternally [206-208, 223] and their proteins

appear be quite stable in the γ-secretase complex [38, 224-226]. The full-length PSENs have a

half-life of 1.5 h, but once cleaved their NTF and CTF half-life is approximately 24 h [38].

Therefore it is possible that sufficient zebrafish Psen protein remains at 24 hours to allow normal

mitochondrion-ER apposition.

160

Morpholinos reportedly remain effective through 50 hpf in zebrafish embryos [227]. Therefore in

order to ensure that negligible or no Psen protein remains at 24 hours, further work will be to

repeat the experiment and observe if changes in apposition are observable at a later time in

development, such as 50 hpf. Using this later time point would ensure that degradation of any

Psen NTF and CTF present at 24 hpf has already occurred. It would also be interesting to study

the effect hypoxia would have on ER-mitochondrial apposition, as hypoxia has been shown to

increase expression of zebrafish bace1, psen1, psen2, and the paralogues of human APP, appa

and appb [123]. Besides hypoxia, certain chemical compounds (such as the former candidate AD

therapeutic drug Latrepirdine) should also be tested to determine if their mechanisms involve

modifying the cross-talk between ER and mitochondria via increased or decreased apposition.

The MAM is enriched in various proteins [228] including proteins with links to AD. The MAM

contains the other components of the γ-secretase complex, as well as its substrate APP [145]. We

have found that the sorting receptor SORL1 is present in the MAM. SORL1 interacts with APP

[111] and either regulates its release from the TGN [112] or assists retrograde transport away

from endosomes, reducing Aβ production [113]. SORL1 levels can affect Aβ production [111] as

reducing SORL1 consequently increases Aβ levels. The YENPTY domain in APP [229] and the

extracellular cluster containing 11 complement-type repeat (CR) domains are important for

binding [230], with mutations to these domains resulting in loss of APP-SORL1 complex

formation.

161

The discovery of APP in the MAM by Area Gomez et al. [145] along with our discovery of

SORL1 in MAM suggests that the MAM is a potential site of initial contact between these two

proteins. If this is the case, it would be interesting to know what effects perturbation of the MAM

would have on APP trafficking to the TGN, endocytic pathway and cell surface. Although

speculative, if MAM dysfunction leads to less binding of SORL1 to APP, then this might result

in increased Aβ production.

The role of autophagy in AD continues to be investigated, but remains unclear. Autophagy

influences secretion of Aβ to the extracellular space, directly affecting Aβ plaque formation

[231]. Researchers have suggested that autophagy may be modulated to enhance β-amyloid

clearance [232]. Therefore attention has been given to finding an effective and safe method to

increase the induction of autophagy and reduce the accumulation of abnormal proteins that cause

neuronal dysfunction. In our work we have shown that the antihistamine drug Latrepirdine can

induce autophagy in zebrafish larvae. In previous studies, there have been contradictory findings

regarding the clinical efficacy of Latrepirdine in patients with AD [233]. Although initially

showing clinical benefits in a pilot study [234], later trials did not indicate any benefits in using

Latrepirdine (discussed in [235]). However, our findings in this thesis and work done in yeast

have indicated that Latrepirdine induces autophagy [220]. In animal models, Latrepirdine appears

to have neuroprotective properties and improves cognition and behaviour [234, 236]. It has also

been shown to be involved in mitochondrial function and calcium homeostasis [237]. Further

meta-analysis of the combined Latrepirdine trials indicates a potential benefit of Latrepirdine on

neuropsychiatric symptoms in AD [233]. Understanding the mechanism for the beneficial use of

Latrepirdine is crucial, and our work indicates that the zebrafish is an advantageous model for the

162

further study of autophagy in AD. It would be interesting to know if Latrepirdine affects MAM

function, as intriguingly, the initial site of formation of the autophagosomes is at the MAM [192].

Autophagy modulation as a potential therapy for AD has been gaining interest among

researchers. However, autophagy modulation may increase extracellular secretion of Aβ, which

in turn could affect plaque formation, although decreasing intracellular Aβ may be beneficial in

reducing intracellular toxicity-induced neurodegeneration. Also of concern is the finding by Yu et

al. that autophagy partakes in Aβ metabolism [238]. Their study showed that γ-secretase activity,

along with PSEN1, NCSTN, APP, βAPP-CTF and Aβ peptides are enriched in autophagic

vacuoles (AVs). In a separate study, Tian et al. discovered that the adaptor complex

AP2/PICALM recruits APP-CTF to autophagosomes [239]. Although believed to result in

degradation, this movement of APP-CTF to AVs could result in more Aβ production instead.

Therefore the induction of autophagy as a form of treatment must be studied carefully, and would

likely require the additional use of Aβ modifying drugs in conjunction.

The MAM as an important factor in both SAD and FAD?

The MAM is important for numerous cellular functions, many of which are dysfunctional in

individuals with AD. For instance the MAM is involved in cholesterol metabolism, which is

altered in AD [130, 131]. The enzyme acyl-coenzyme A: cholesterol acyltransferase-1 (ACAT1)

catalyzes cholesteryl ester (CE) formation from free cholesterol (FC) and maintains their

equilibrium, making it important for cholesterol homeostasis (reviewed in [240]). Interestingly

the MAM is a platform for the transfer and synthesis of lipids, and is highly enriched in ACAT1

163

[150]. A reduction in CE due to decreased ACAT1 expression leads to a decline in Aβ

production, although the mechanistic details of this occurrence are still uncertain [241]. Likewise

an increase in CE levels upregulates generation of Aβ, and is due to alteration in the rate of Aβ

generation [242]. The activity of the α-, β- and γ-secretases appears to be affected when ACAT1

levels are reduced since Aβ, APP-C99 and also APP-C83 levels were decreased [241]. The link

between MAM and the PSENs is that the MAM is a known subcellular location of non-truncated

PRESENILINs [145]. Cholesteryl ester and phospholipid synthesis is elevated in MEFs lacking

PSENs and in FAD and also in SAD patient fibroblasts [158] although the mechanisms remain

unclear. This upregulation would likely produce more Aβ. The changes in the profile of Aβ

lengths may also be a consequence of changes in the lipid constitution of the MAM itself and not

necessarily due to the effect of PSEN mutations on cleavage activity.

Aβ itself appears capable of causing mitochondrial dysfunction. Aβ can be imported into

mitochondria using the translocase of the outer membrane (TOM) machinery [243]. Evidence

suggests accumulation of Aβ in mitochondria affects bioenergetics, with decreases in ATP

production and increased ROS production (reviewed in [244, 245]). Normal aerobic respiration

produces ROS by mitochondria during electron transport in the oxidative phosphorylation chain.

Under normal conditions these ROS are mainly sequestered within mitochondria (reviewed in

[246]). However some leakage of ROS does occur and causes damage to the surroundings. The

presence of antioxidant mechanisms are meant to reduce the level of damage by sequestering

ROS [247]. However the increased ROS production due to impaired mitochondrial function

would reduce their effectiveness and likely cause increased oxidative damage-induced neuronal

164

death. Therefore it is possible that a dysfunction of the MAM could ultimately result in the

increased ROS and damaged mitochondria previously observed in AD patients, at least in SAD.

Certain neurodegenerative disorders including AD appear to share the common feature of protein

aggregation within the CNS (reviewed in [248]). Metabolic energy is required for proper

disulfide bond formation, and a lack of ATP might result in misfolding [249, 250]. Interestingly

the MAM is rich in oxidoreductases, and is the site of disulfide bond folding (reviewed in [146]),

suggesting that MAM disruption could also be a cause of protein aggregation.

Ca2+ dysregulation has been noted in AD patients (reviewed in [251]). Not surprisingly, the

MAM is the location of several Ca2+ receptors (reviewed in [228]) and plays a role in regulating

Ca2+ release to mitochondria. Interestingly, the zymogen forms of the PSEN proteins act as Ca2+

leak channels in the ER, and FAD mutations appear to interfere with this ability [42]. Several

functions of mitochondria are dependent upon Ca2+ levels in the cell. For instance, mitochondrial

motility is dependent upon cytosolic concentrations of Ca2+ ([Ca2+]c), as mitochondria are

retained at sites of [Ca2+]c elevations [252]. This suggests that a dysfunction of Ca2+ release at

the MAM might redirect mitochondria away from areas that require them for energy production,

such as the synapses of neurons.

165

One of the important roles of Ca2+ is its role in mitochondrial metabolism (reviewed in [253]).

However, too close an association between mitochondria and MAM could result in toxic

mitochondria Ca2+ overload [254], with the activation of apoptotic cell death. Work done by

Area-Gomez et al. has found the number of ER-mitochondrial contacts was increased in MEFs

lacking PSENs and in both FAD and SAD patient fibroblasts with increased apposition lengths

between ER and mitochondria [158]. Furthermore Aβ exposure increases the number of contact

points between ER and mitochondria [255]. In terms of energy requirement, the brain is a high

energy organ and is highly dependent on ATP production by mitochondria, which are abundant at

synapses of neurons [256], but the AD brain shows deficits in brain metabolism [257], with

evidence that dysfunction of mitochondria is a likely cause.

As was mentioned earlier, autophagy is linked to the MAM as it is the initiation site of

autophagosome formation. Moreover, PSEN1 also partakes in the pathway, as its presence is

required for lysosomal acidification and proteolysis during autophagy [41]. Taken together, it is

noticeable that the multitudes of abnormal features observed in AD are the same processes which

involve the MAM. Not only are these features observed in SAD, many are also observed in FAD.

Added to this is the fact that the FAD-related proteins, PSEN1 and PSEN2 are enriched in the

MAM, and that other risk proteins such as the SAD-associated SORL1 are also localized there. It

is becoming more apparent that the MAM plays a central role in AD pathogenesis, and that the

work done thus far regarding MAM has given us some small insight into how AD symptoms are

linked together. Many of the pathways that MAM plays a role in (autophagy, Ca2+, lipid and

cholesterol dysregulation, mitochondrial damage and energy deficiency, protein misfolding) are

interconnected such that disruption to one of the MAM’s functions has the potency to result in

166

the disturbance of the MAM’s other tasks. It is likely this cycle of interference within various

cellular pathways ultimately culminates in neuronal cell loss and death. How the mutations in the

PSENs affect MAM-mitochondrial apposition remains unclear, but it is hoped that the future

work to validate the use of the zebrafish animal model as a useful model in studying apposition

will reveal its successful use to further studying the link between MAM and the PSENs.

In summary, AD is a complex disease with numerous genetic and environmental influences. To

fully understand the mechanisms of AD pathogenesis, it is important to identify the subcellular

localization of proteins implicated in AD. The discovery of the MAM as a significant site of

enrichment of the PRESENILINS and other AD-risk related proteins such as SORL1, as well as

its role in features known to be perturbed in AD will likely have a large impact on our

understanding of AD pathogenesis.

167

REFERENCES

1. Alzheimer's, A., 2010 Alzheimer's disease facts and figures. Alzheimers Dement, 2010. 6(2): p. 158-94.

2. Ferri, C.P., et al., Global prevalence of dementia: a Delphi consensus study. Lancet, 2005. 366(9503): p. 2112-7.

3. Voisin, T. and B. Vellas, Diagnosis and treatment of patients with severe Alzheimer's disease. Drugs & aging, 2009. 26(2): p. 135-44.

4. Grundke-Iqbal, I., et al., Abnormal phosphorylation of the microtubule-associated protein tau (tau) in Alzheimer cytoskeletal pathology. Proceedings of the National Academy of Sciences of the United States of America, 1986. 83(13): p. 4913-4917.

5. Pettegrew, J.W., et al., Brain membrane phospholipid alterations in Alzheimer's disease. Neurochemical research, 2001. 26(7): p. 771-82.

6. Bezprozvanny, I. and M.P. Mattson, Neuronal calcium mishandling and the pathogenesis of Alzheimer's disease. Trends in neurosciences, 2008. 31(9): p. 454-63.

7. Muirhead, K.E., et al., The consequences of mitochondrial amyloid beta-peptide in Alzheimer's disease. The Biochemical journal, 2010. 426(3): p. 255-70.

8. Bekris, L.M., et al., Genetics of Alzheimer disease. Journal of geriatric psychiatry and neurology, 2010. 23(4): p. 213-27.

9. Campion, D., et al., Early-onset autosomal dominant Alzheimer disease: prevalence, genetic heterogeneity, and mutation spectrum. American journal of human genetics, 1999. 65(3): p. 664-670.

10. Hardy, J.A. and G.A. Higgins, Alzheimer's disease: the amyloid cascade hypothesis. Science, 1992. 256(5054): p. 184-5.

11. Axelsen, P.H., H. Komatsu, and I.V.J. Murray, Oxidative Stress and Cell Membranes in the Pathogenesis of Alzheimer's Disease. Physiology, 2011. 26(1): p. 54-69.

12. Pratico, D., et al., Increased lipid peroxidation precedes amyloid plaque formation in an animal model of Alzheimer amyloidosis. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2001. 21(12): p. 4183-7.

13. Hensley, K., et al., Brain regional correspondence between Alzheimer's disease histopathology and biomarkers of protein oxidation. Journal of neurochemistry, 1995. 65(5): p. 2146-56.

14. Jeandel, C., et al., Lipid peroxidation and free radical scavengers in Alzheimer's disease. Gerontology, 1989. 35(5-6): p. 275-82.

15. Zaman, Z., et al., Plasma concentrations of vitamins A and E and carotenoids in Alzheimer's disease. Age Ageing, 1992. 21(2): p. 91-4.

16. Kontush, A., N. Donarski, and U. Beisiegel, Resistance of human cerebrospinal fluid to in vitro oxidation is directly related to its amyloid-beta content. Free radical research, 2001. 35(5): p. 507-17.

17. Zou, K., et al., A novel function of monomeric amyloid beta-protein serving as an antioxidant molecule against metal-induced oxidative damage. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2002. 22(12): p. 4833-41.

18. Dineley, K.T., et al., Beta-amyloid activates the mitogen-activated protein kinase cascade via hippocampal alpha7 nicotinic acetylcholine receptors: In vitro and in vivo mechanisms related

168

to Alzheimer's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2001. 21(12): p. 4125-33.

19. Wang, Q., et al., Block of long-term potentiation by naturally secreted and synthetic amyloid beta-peptide in hippocampal slices is mediated via activation of the kinases c-Jun N-terminal kinase, cyclin-dependent kinase 5, and p38 mitogen-activated protein kinase as well as metabotropic glutamate receptor type 5. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2004. 24(13): p. 3370-8.

20. Huang, H.M. and G.E. Gibson, Altered beta-adrenergic receptor-stimulated cAMP formation in cultured skin fibroblasts from Alzheimer donors. Journal of Biological Chemistry, 1993. 268(20): p. 14616-14621.

21. Townsend, M., T. Mehta, and D.J. Selkoe, Soluble Abeta inhibits specific signal transduction cascades common to the insulin receptor pathway. The Journal of biological chemistry, 2007. 282(46): p. 33305-12.

22. Dovey, H.F., et al., Cells with a familial Alzheimer's disease mutation produce authentic beta-peptide. Neuroreport, 1993. 4(8): p. 1039-42.

23. Asami-Odaka, A., et al., Long amyloid beta-protein secreted from wild-type human neuroblastoma IMR-32 cells. Biochemistry, 1995. 34(32): p. 10272-8.

24. Thinakaran, G. and E.H. Koo, Amyloid precursor protein trafficking, processing, and function. The Journal of biological chemistry, 2008. 283(44): p. 29615-9.

25. Krishnaswamy, S., et al., The structure and function of Alzheimer's gamma secretase enzyme complex. Critical reviews in clinical laboratory sciences, 2009. 46(5-6): p. 282-301.

26. Schroeter, E.H., J.A. Kisslinger, and R. Kopan, Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature, 1998. 393(6683): p. 382-6.

27. Artavanis-Tsakonas, S., K. Matsuno, and M.E. Fortini, Notch signaling. Science, 1995. 268(5208): p. 225-32.

28. Struhl, G. and I. Greenwald, Presenilin-mediated transmembrane cleavage is required for Notch signal transduction in Drosophila. Proceedings of the National Academy of Sciences of the United States of America, 2001. 98(1): p. 229-34.

29. Hemming, M.L., et al., Proteomic profiling of gamma-secretase substrates and mapping of substrate requirements. PLoS Biol, 2008. 6(10): p. e257.

30. Haapasalo, A. and D.M. Kovacs, The many substrates of presenilin/gamma-secretase. Journal of Alzheimer's disease : JAD, 2011. 25(1): p. 3-28.

31. McCarthy, J.V., C. Twomey, and P. Wujek, Presenilin-dependent regulated intramembrane proteolysis and gamma-secretase activity. Cellular and molecular life sciences : CMLS, 2009. 66(9): p. 1534-55.

32. Sato, T., et al., Active gamma-secretase complexes contain only one of each component. Journal of Biological Chemistry, 2007. 282(47): p. 33985-33993.

33. Levy-Lahad, E., et al., Candidate gene for the chromosome 1 familial Alzheimer's disease locus. Science, 1995. 269(5226): p. 973-7.

34. De Jonghe, C., et al., Aberrant splicing in the presenilin-1 intron 4 mutation causes presenile Alzheimer's disease by increased Abeta42 secretion. Human molecular genetics, 1999. 8(8): p. 1529-40.

35. Laudon, H., et al., A nine-transmembrane domain topology for presenilin 1. The Journal of biological chemistry, 2005. 280(42): p. 35352-60.

36. Fukumori, A., et al., Three-Amino Acid Spacing of Presenilin Endoproteolysis Suggests a General Stepwise Cleavage of gamma-Secretase-Mediated Intramembrane Proteolysis. Journal of Neuroscience, 2010. 30(23): p. 7853-7862.

169

37. Mann, D.M., et al., Cases of Alzheimer's disease due to deletion of exon 9 of the presenilin-1 gene show an unusual but characteristic beta-amyloid pathology known as 'cotton wool' plaques. Neuropathology and applied neurobiology, 2001. 27(3): p. 189-96.

38. Ratovitski, T., et al., Endoproteolytic processing and stabilization of wild-type and mutant presenilin. The Journal of biological chemistry, 1997. 272(39): p. 24536-41.

39. Wolfe, M.S., et al., Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and gamma-secretase activity. Nature, 1999. 398(6727): p. 513-7.

40. Kimberly, W.T., et al., The transmembrane aspartates in presenilin 1 and 2 are obligatory for gamma-secretase activity and amyloid beta-protein generation. The Journal of biological chemistry, 2000. 275(5): p. 3173-8.

41. Lee, J.H., et al., Lysosomal proteolysis and autophagy require presenilin 1 and are disrupted by Alzheimer-related PS1 mutations. Cell, 2010. 141(7): p. 1146-58.

42. Tu, H., et al., Presenilins form ER Ca2+ leak channels, a function disrupted by familial Alzheimer's disease-linked mutations. Cell, 2006. 126(5): p. 981-93.

43. Zhang, H., et al., Role of presenilins in neuronal calcium homeostasis. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2010. 30(25): p. 8566-80.

44. Yu, G., et al., Nicastrin modulates presenilin-mediated notch/glp-1 signal transduction and betaAPP processing. Nature, 2000. 407(6800): p. 48-54.

45. Yang, D.S., et al., Mature glycosylation and trafficking of nicastrin modulate its binding to presenilins. The Journal of biological chemistry, 2002. 277(31): p. 28135-42.

46. Chavez-Gutierrez, L., et al., Glu(332) in the Nicastrin ectodomain is essential for gamma-secretase complex maturation but not for its activity. The Journal of biological chemistry, 2008. 283(29): p. 20096-105.

47. Zhao, G., et al., Gamma-secretase composed of PS1/Pen2/Aph1a can cleave notch and amyloid precursor protein in the absence of nicastrin. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2010. 30(5): p. 1648-56.

48. Futai, E., S. Yagishita, and S. Ishiura, Nicastrin Is Dispensable for γ-Secretase Protease Activity in the Presence of Specific Presenilin Mutations. The Journal of biological chemistry, 2009. 284(19): p. 13013-13022.

49. Shah, S., et al., Nicastrin functions as a gamma-secretase-substrate receptor. Cell, 2005. 122(3): p. 435-47.

50. Zhang, X., et al., Identification of a tetratricopeptide repeat-like domain in the nicastrin subunit of gamma-secretase using synthetic antibodies. Proceedings of the National Academy of Sciences of the United States of America, 2012. 109(22): p. 8534-9.

51. Xie, T., et al., Crystal structure of the gamma-secretase component nicastrin. Proceedings of the National Academy of Sciences of the United States of America, 2014. 111(37): p. 13349-54.

52. Fortna, R.R., et al., Membrane topology and nicastrin-enhanced endoproteolysis of APH-1, a component of the gamma-secretase complex. The Journal of biological chemistry, 2004. 279(5): p. 3685-93.

53. Shirotani, K., et al., Identification of distinct gamma-secretase complexes with different APH-1 variants. The Journal of biological chemistry, 2004. 279(40): p. 41340-5.

54. Serneels, L., et al., Differential contribution of the three Aph1 genes to gamma-secretase activity in vivo. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(5): p. 1719-24.

55. Ma, G., et al., APH-1a is the principal mammalian APH-1 isoform present in gamma-secretase complexes during embryonic development. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2005. 25(1): p. 192-8.

170

56. Capell, A., et al., Gamma-secretase complex assembly within the early secretory pathway. The Journal of biological chemistry, 2005. 280(8): p. 6471-8.

57. Crystal, A.S., et al., Membrane topology of gamma-secretase component PEN-2. The Journal of biological chemistry, 2003. 278(22): p. 20117-23.

58. Luo, W.J., et al., PEN-2 and APH-1 coordinately regulate proteolytic processing of presenilin 1. The Journal of biological chemistry, 2003. 278(10): p. 7850-4.

59. Prokop, S., et al., Requirement of PEN-2 for stabilization of the presenilin N-/C-terminal fragment heterodimer within the gamma-secretase complex. The Journal of biological chemistry, 2004. 279(22): p. 23255-61.

60. Steiner, H., E. Winkler, and C. Haass, Chemical cross-linking provides a model of the gamma-secretase complex subunit architecture and evidence for close proximity of the C-terminal fragment of presenilin with APH-1. The Journal of biological chemistry, 2008. 283(50): p. 34677-86.

61. Kaether, C., et al., The presenilin C-terminus is required for ER-retention, nicastrin-binding and gamma-secretase activity. The EMBO journal, 2004. 23(24): p. 4738-48.

62. Lu, P., et al., Three-dimensional structure of human gamma-secretase. Nature, 2014. 512(7513): p. 166-70.

63. Li, X., et al., Structure of a presenilin family intramembrane aspartate protease. Nature, 2013. 493(7430): p. 56-61.

64. De Strooper, B. and W. Annaert, Novel research horizons for presenilins and gamma-secretases in cell biology and disease. Annual review of cell and developmental biology, 2010. 26: p. 235-60.

65. Goate, A., et al., Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature, 1991. 349(6311): p. 704-6.

66. Sherrington, R., et al., Cloning of a gene bearing missense mutations in early-onset familial Alzheimer's disease. Nature, 1995. 375(6534): p. 754-60.

67. Dermaut, B., et al., A novel presenilin 1 mutation associated with Pick's disease but not beta-amyloid plaques. Annals of neurology, 2004. 55(5): p. 617-26.

68. Li, D.X., et al., Mutations of presenilin genes in dilated cardiomyopathy and heart failure. American journal of human genetics, 2006. 79(6): p. 1030-1039.

69. Fitzsimmons, J.S. and P.R. Guilbert, A family study of hidradenitis suppurativa. Journal of medical genetics, 1985. 22(5): p. 367-73.

70. De Jonghe, C., et al., Pathogenic APP mutations near the γ-secretase cleavage site differentially affect Aβ secretion and APP C-terminal fragment stability. Human molecular genetics, 2001. 10(16): p. 1665-1671.

71. Nornes, S., et al., Interference with splicing of Presenilin transcripts has potent dominant negative effects on Presenilin activity. Human molecular genetics, 2008. 17(3): p. 402-12.

72. Kwok, J.B., et al., Presenilin-1 mutation L271V results in altered exon 8 splicing and Alzheimer's disease with non-cored plaques and no neuritic dystrophy. The Journal of biological chemistry, 2003. 278(9): p. 6748-54.

73. Perez-Tur, J., et al., A mutation in Alzheimer's disease destroying a splice acceptor site in the presenilin-1 gene. Neuroreport, 1995. 7(1): p. 297-301.

74. Hiltunen, M., et al., Identification of a novel 4.6-kb genomic deletion in presenilin-1 gene which results in exclusion of exon 9 in a Finnish early onset Alzheimer's disease family: an Alu core sequence-stimulated recombination? European journal of human genetics : EJHG, 2000. 8(4): p. 259-66.

75. Seelaar, H., et al., Clinical, genetic and pathological heterogeneity of frontotemporal dementia: a review. Journal of neurology, neurosurgery, and psychiatry, 2011. 82(5): p. 476-86.

171

76. Wang, B., et al., Gamma-secretase gene mutations in familial acne inversa. Science, 2010. 330(6007): p. 1065.

77. Corey, D.R. and J.M. Abrams, Morpholino antisense oligonucleotides: tools for investigating vertebrate development. Genome biology, 2001. 2(5): p. REVIEWS1015.

78. Nornes, S., et al., Independent and cooperative action of Psen2 with Psen1 in zebrafish embryos. Experimental cell research, 2009. 315(16): p. 2791-801.

79. Rechards, M., et al., Presenilin-1 exists in both pre- and post-Golgi compartments and recycles via COPI-coated membranes. Traffic, 2003. 4(8): p. 553-65.

80. Tarassishin, L., et al., Processing of Notch and amyloid precursor protein by gamma-secretase is spatially distinct. Proceedings of the National Academy of Sciences of the United States of America, 2004. 101(49): p. 17050-5.

81. Kopan, R. and M.X. Ilagan, The canonical Notch signaling pathway: unfolding the activation mechanism. Cell, 2009. 137(2): p. 216-33.

82. Heyman, A., et al., Alzheimer's disease: A study of epidemiological aspects. Annals of neurology, 1984. 15(4): p. 335-341.

83. Heston, L.L., et al., Dementia of the alzheimer type: Clinical genetics, natural history, and associated conditions. Archives of general psychiatry, 1981. 38(10): p. 1085-1090.

84. Merched, A., et al., Decreased high-density lipoprotein cholesterol and serum apolipoprotein AI concentrations are highly correlated with the severity of Alzheimer's disease. Neurobiology of aging, 2000. 21(1): p. 27-30.

85. Kuo, Y.M., et al., Elevated low-density lipoprotein in Alzheimer's disease correlates with brain abeta 1-42 levels. Biochemical and biophysical research communications, 1998. 252(3): p. 711-5.

86. Gustafson, D., et al., An 18-year follow-up of overweight and risk of Alzheimer disease. Archives of internal medicine, 2003. 163(13): p. 1524-8.

87. Luchsinger, J.A., et al., Diabetes mellitus and risk of Alzheimer's disease and dementia with stroke in a multiethnic cohort. American journal of epidemiology, 2001. 154(7): p. 635-41.

88. Reed, B., et al., Associations between serum cholesterol levels and cerebral amyloidosis. JAMA Neurol, 2014. 71(2): p. 195-200.

89. Refolo, L.M., et al., Hypercholesterolemia accelerates the Alzheimer's amyloid pathology in a transgenic mouse model. Neurobiology of disease, 2000. 7(4): p. 321-31.

90. Shie, F.S., et al., Diet-induced hypercholesterolemia enhances brain A beta accumulation in transgenic mice. Neuroreport, 2002. 13(4): p. 455-9.

91. Sparks, D.L., et al., Alterations of Alzheimer's disease in the cholesterol-fed rabbit, including vascular inflammation. Preliminary observations. Annals of the New York Academy of Sciences, 2000. 903: p. 335-44.

92. Craft, S., The role of metabolic disorders in Alzheimer disease and vascular dementia: two roads converged. Archives of neurology, 2009. 66(3): p. 300-5.

93. Sridhar, G.R., G. Lakshmi, and G. Nagamani, Emerging links between type 2 diabetes and Alzheimer’s disease. World Journal of Diabetes, 2015. 6(5): p. 744-751.

94. Kawamura, T., T. Umemura, and N. Hotta, Cognitive impairment in diabetic patients: Can diabetic control prevent cognitive decline? Journal of Diabetes Investigation, 2012. 3(5): p. 413-423.

95. Qiu, W.Q., et al., Insulin-degrading enzyme regulates extracellular levels of amyloid beta-protein by degradation. The Journal of biological chemistry, 1998. 273(49): p. 32730-8.

96. Poole, S., et al., Determining the presence of periodontopathic virulence factors in short-term postmortem Alzheimer's disease brain tissue. Journal of Alzheimer's disease : JAD, 2013. 36(4): p. 665-77.

172

97. Ayton, S., et al., Ferritin levels in the cerebrospinal fluid predict Alzheimer/'s disease outcomes and are regulated by APOE. Nat Commun, 2015. 6.

98. Suwalsky, M., et al., Effects of Al(III) speciation on cell membranes and molecular models. Coordination Chemistry Reviews, 2002. 228(2): p. 285-295.

99. Harold, D., et al., Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nature genetics, 2009. 41(10): p. 1088-U61.

100. Strittmatter, W.J., et al., Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proceedings of the National Academy of Sciences of the United States of America, 1993. 90(5): p. 1977-81.

101. Bjorkhem, I. and S. Meaney, Brain cholesterol: long secret life behind a barrier. Arteriosclerosis, thrombosis, and vascular biology, 2004. 24(5): p. 806-15.

102. Castellano, J.M., et al., Human apoE isoforms differentially regulate brain amyloid-beta peptide clearance. Science translational medicine, 2011. 3(89): p. 89ra57.

103. Holtzman, D.M., et al., Apolipoprotein E isoform-dependent amyloid deposition and neuritic degeneration in a mouse model of Alzheimer's disease. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(6): p. 2892-2897.

104. Hollingworth, P., et al., Alzheimer's disease genetics: current knowledge and future challenges. International Journal of Geriatric Psychiatry, 2011. 26(8): p. 793-802.

105. Martins, I.J., et al., Apolipoprotein E, cholesterol metabolism, diabetes, and the convergence of risk factors for Alzheimer's disease and cardiovascular disease. Molecular psychiatry, 2006. 11(8): p. 721-36.

106. Meng, Y., et al., Association between SORL1 and Alzheimer's disease in a genome-wide study. Neuroreport, 2007. 18(17): p. 1761-4.

107. Miyashita, A., et al., SORL1 is genetically associated with late-onset Alzheimer's disease in Japanese, Koreans and Caucasians. PloS one, 2013. 8(4): p. e58618.

108. Wen, Y., et al., SORL1 is genetically associated with neuropathologically characterized late-onset Alzheimer's disease. Journal of Alzheimer's disease : JAD, 2013. 35(2): p. 387-94.

109. Motoi, Y., et al., Neuronal localization of a novel mosaic apolipoprotein E receptor, LR11, in rat and human brain. Brain research, 1999. 833(2): p. 209-15.

110. Jacobsen, L., et al., Activation and functional characterization of the mosaic receptor SorLA/LR11. The Journal of biological chemistry, 2001. 276(25): p. 22788-96.

111. Andersen, O.M., et al., Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(38): p. 13461-13466.

112. Schmidt, V., et al., SorLA/LR11 Regulates Processing of Amyloid Precursor Protein via Interaction with Adaptors GGA and PACS-1. Journal of Biological Chemistry, 2007. 282(45): p. 32956-32964.

113. Lane, R.F., et al., Protein kinase C and rho activated coiled coil protein kinase 2 (ROCK2) modulate Alzheimer's APP metabolism and phosphorylation of the Vps10-domain protein, SorL1. Molecular neurodegeneration, 2010. 5: p. 62.

114. Scherzer, C.R., et al., Loss of apolipoprotein E receptor LR11 in Alzheimer disease. Archives of neurology, 2004. 61(8): p. 1200-5.

115. Sager, K.L., et al., Neuronal LR11/sorLA expression is reduced in mild cognitive impairment. Annals of neurology, 2007. 62(6): p. 640-7.

116. Zhu, X., et al., Vascular oxidative stress in Alzheimer disease. Journal of the neurological sciences, 2007. 257(1-2): p. 240-6.

173

117. Gama Sosa, M.A., et al., Age-Related Vascular Pathology in Transgenic Mice Expressing Presenilin 1-Associated Familial Alzheimer's Disease Mutations. The American journal of pathology, 2010. 176(1): p. 353-368.

118. Zhang, X., et al., Hypoxia-inducible factor 1alpha (HIF-1alpha)-mediated hypoxia increases BACE1 expression and beta-amyloid generation. The Journal of biological chemistry, 2007. 282(15): p. 10873-80.

119. Mohuczy, D., K. Qian, and M.I. Phillips, Presenilins in the heart: presenilin-2 expression is increased by low glucose and by hypoxia in cardiac cells. Regulatory peptides, 2002. 110(1): p. 1-7.

120. Tamagno, E., et al., Oxidative stress activates a positive feedback between the γ- and β-secretase cleavages of the β-amyloid precursor protein. Journal of neurochemistry, 2008. 104(3): p. 683-695.

121. Manabe, T., et al., HMGA1a: sequence-specific RNA-binding factor causing sporadic Alzheimer's disease-linked exon skipping of presenilin-2 pre-mRNA. Genes to cells : devoted to molecular & cellular mechanisms, 2007. 12(10): p. 1179-91.

122. Moussavi Nik, S.H., M. Newman, and M. Lardelli, The response of HMGA1 to changes in oxygen availability is evolutionarily conserved. Experimental cell research, 2011. 317(11): p. 1503-12.

123. Moussavi Nik, S.H., et al., The BACE1-PSEN-AbetaPP regulatory axis has an ancient role in response to low oxygen/oxidative stress. Journal of Alzheimer's disease : JAD, 2012. 28(3): p. 515-30.

124. Klunk, W.E., et al., Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Annals of neurology, 2004. 55(3): p. 306-319.

125. Fleisher, A.S., et al., USing positron emission tomography and florbetapir f 18 to image cortical amyloid in patients with mild cognitive impairment or dementia due to alzheimer disease. Archives of neurology, 2011. 68(11): p. 1404-1411.

126. Morris, G.P., I.A. Clark, and B. Vissel, Inconsistencies and controversies surrounding the amyloid hypothesis of Alzheimer's disease. Acta Neuropathol Commun, 2014. 2: p. 135.

127. Balasubramanian, A.B., et al., Alzheimer disease pathology and longitudinal cognitive performance in the oldest-old with no dementia. Neurology, 2012. 79(9): p. 915-921.

128. Price, J.L., et al., Neuropathology of nondemented aging: Presumptive evidence for preclinical Alzheimer disease. Neurobiology of aging, 2009. 30(7): p. 1026-1036.

129. Zigman, W.B., et al., Alzheimer’s Disease in Adults with Down Syndrome. International review of research in mental retardation, 2008. 36: p. 103-145.

130. Pappolla, M.A., et al., Mild hypercholesterolemia is an early risk factor for the development of Alzheimer amyloid pathology. Neurology, 2003. 61(2): p. 199-205.

131. Solomon, A., et al., Midlife Serum Cholesterol and Increased Risk of Alzheimer's and Vascular Dementia Three Decades Later. Dementia and geriatric cognitive disorders, 2009. 28(1): p. 75-80.

132. Berridge, M.J., Calcium hypothesis of Alzheimer's disease. Pflugers Archiv : European journal of physiology, 2010. 459(3): p. 441-9.

133. Swerdlow, R.H., J.M. Burns, and S.M. Khan, The Alzheimer's disease mitochondrial cascade hypothesis. Journal of Alzheimer's disease : JAD, 2010. 20 Suppl 2: p. S265-79.

134. Gong, C.X., et al., Impaired brain glucose metabolism leads to Alzheimer neurofibrillary degeneration through a decrease in tau O-GlcNAcylation. Journal of Alzheimer's disease : JAD, 2006. 9(1): p. 1-12.

135. Ringheim, G.E. and A.M. Szczepanik, Brain inflammation, cholesterol, and glutamate as interconnected participants in the pathology of Alzheimer's disease. Current pharmaceutical design, 2006. 12(6): p. 719-38.

174

136. Takashima, A., et al., Presenilin 1 associates with glycogen synthase kinase-3beta and its substrate tau. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(16): p. 9637-41.

137. Pasternak, S.H., Presenilin-1, Nicastrin, Amyloid Precursor Protein, and -Secretase Activity Are Co-localized in the Lysosomal Membrane. Journal of Biological Chemistry, 2003. 278(29): p. 26687-26694.

138. Huse, J.T., et al., Maturation and Endosomal Targeting of β-Site Amyloid Precursor Protein-cleaving Enzyme. Journal of Biological Chemistry, 2000. 275(43): p. 33729-33737.

139. Koo, E.H. and S.L. Squazzo, Evidence that production and release of amyloid beta-protein involves the endocytic pathway. The Journal of biological chemistry, 1994. 269(26): p. 17386-9.

140. Huse, J.T., et al., Beta-secretase processing in the trans-Golgi network preferentially generates truncated amyloid species that accumulate in Alzheimer's disease brain. The Journal of biological chemistry, 2002. 277(18): p. 16278-84.

141. Annaert, W.G., et al., Presenilin 1 controls gamma-secretase processing of amyloid precursor protein in pre-golgi compartments of hippocampal neurons. The Journal of cell biology, 1999. 147(2): p. 277-94.

142. Vetrivel, K.S., et al., Association of gamma-secretase with lipid rafts in post-Golgi and endosome membranes. The Journal of biological chemistry, 2004. 279(43): p. 44945-54.

143. Ankarcrona, M. and K. Hultenby, Presenilin-1 is located in rat mitochondria. Biochemical and biophysical research communications, 2002. 295(3): p. 766-70.

144. Li, J., et al., Alzheimer presenilins in the nuclear membrane, interphase kinetochores, and centrosomes suggest a role in chromosome segregation. Cell, 1997. 90(5): p. 917-27.

145. Area-Gomez, E., et al., Presenilins are enriched in endoplasmic reticulum membranes associated with mitochondria. The American journal of pathology, 2009. 175(5): p. 1810-6.

146. Simmen, T., et al., Oxidative protein folding in the endoplasmic reticulum: Tight links to the mitochondria-associated membrane (MAM). Biochimica Et Biophysica Acta-Biomembranes, 2010. 1798(8): p. 1465-1473.

147. Hayashi, T., et al., MAM: more than just a housekeeper. Trends in cell biology, 2009. 19(2): p. 81-8.

148. Stone, S.J. and J.E. Vance, Phosphatidylserine Synthase-1 and -2 Are Localized to Mitochondria-associated Membranes. Journal of Biological Chemistry, 2000. 275(44): p. 34534-34540.

149. Zborowski, J., A. Dygas, and L. Wojtczak, Phosphatidylserine decarboxylase is located on the external side of the inner mitochondrial membrane. FEBS letters, 1983. 157(1): p. 179-82.

150. Rusinol, A.E., et al., A unique mitochondria-associated membrane fraction from rat liver has a high capacity for lipid synthesis and contains pre-Golgi secretory proteins including nascent lipoproteins. The Journal of biological chemistry, 1994. 269(44): p. 27494-502.

151. Benham, A.M., et al., The CXXCXXC motif determines the folding, structure and stability of human Ero1-Lalpha. The EMBO journal, 2000. 19(17): p. 4493-502.

152. Mezghrani, A., et al., Manipulation of oxidative protein folding and PDI redox state in mammalian cells. The EMBO journal, 2001. 20(22): p. 6288-6296.

153. Frand, A.R. and C.A. Kaiser, Ero1p oxidizes protein disulfide isomerase in a pathway for disulfide bond formation in the endoplasmic reticulum. Molecular cell, 1999. 4(4): p. 469-77.

154. Vance, J.E., PHOSPHOLIPID-SYNTHESIS IN A MEMBRANE-FRACTION ASSOCIATED WITH MITOCHONDRIA. Journal of Biological Chemistry, 1990. 265(13): p. 7248-7256.

155. Schon, E.A. and E. Area-Gomez, Is Alzheimer's disease a disorder of mitochondria-associated membranes? Journal of Alzheimer's disease : JAD, 2010. 20 Suppl 2: p. S281-92.

156. Lustbader, J.W., et al., ABAD directly links Abeta to mitochondrial toxicity in Alzheimer's disease. Science, 2004. 304(5669): p. 448-52.

175

157. Zhu, X., et al., Mitochondrial failures in Alzheimer's disease. American journal of Alzheimer's disease and other dementias, 2004. 19(6): p. 345-52.

158. Area-Gomez, E., et al., Upregulated function of mitochondria-associated ER membranes in Alzheimer disease. The EMBO journal, 2012. 31(21): p. 4106-23.

159. He, C. and D.J. Klionsky, Regulation mechanisms and signaling pathways of autophagy. Annu Rev Genet, 2009. 43: p. 67-93.

160. Nakagawa, I., et al., Autophagy defends cells against invading group A Streptococcus. Science, 2004. 306(5698): p. 1037-40.

161. Gutierrez, M.G., et al., Autophagy is a defense mechanism inhibiting BCG and Mycobacterium tuberculosis survival in infected macrophages. Cell, 2004. 119(6): p. 753-66.

162. Li, W.W., J. Li, and J.K. Bao, Microautophagy: lesser-known self-eating. Cellular and molecular life sciences : CMLS, 2012. 69(7): p. 1125-36.

163. Fred Dice, J., Peptide sequences that target cytosolic proteins for lysosomal proteolysis. Trends in biochemical sciences, 1990. 15(8): p. 305-309.

164. Majeski, A.E. and J. Fred Dice, Mechanisms of chaperone-mediated autophagy. The International Journal of Biochemistry & Cell Biology, 2004. 36(12): p. 2435-2444.

165. Shang, L. and X. Wang, AMPK and mTOR coordinate the regulation of Ulk1 and mammalian autophagy initiation. Autophagy, 2011. 7(8): p. 924-6.

166. Di Bartolomeo, S., et al., The dynamic interaction of AMBRA1 with the dynein motor complex regulates mammalian autophagy. The Journal of cell biology, 2010. 191(1): p. 155-68.

167. Polson, H.E., et al., Mammalian Atg18 (WIPI2) localizes to omegasome-anchored phagophores and positively regulates LC3 lipidation. Autophagy, 2010. 6(4): p. 506-22.

168. Obara, K., et al., The Atg18-Atg2 Complex Is Recruited to Autophagic Membranes via Phosphatidylinositol 3-Phosphate and Exerts an Essential Function. The Journal of biological chemistry, 2008. 283(35): p. 23972-23980.

169. Zhong, Y., et al., Distinct regulation of autophagic activity by Atg14L and Rubicon associated with Beclin 1-phosphatidylinositol-3-kinase complex. Nature cell biology, 2009. 11(4): p. 468-76.

170. Matsunaga, K., et al., Two Beclin 1-binding proteins, Atg14L and Rubicon, reciprocally regulate autophagy at different stages. Nature cell biology, 2009. 11(4): p. 385-96.

171. Liang, C., et al., Autophagic and tumour suppressor activity of a novel Beclin1-binding protein UVRAG. Nature cell biology, 2006. 8(7): p. 688-99.

172. Takahashi, Y., et al., Bif-1 interacts with Beclin 1 through UVRAG and regulates autophagy and tumorigenesis. Nature cell biology, 2007. 9(10): p. 1142-51.

173. Takahashi, Y., C.L. Meyerkord, and H.G. Wang, Bif-1/endophilin B1: a candidate for crescent driving force in autophagy. Cell death and differentiation, 2009. 16(7): p. 947-55.

174. Yang, Z. and D.J. Klionsky, Mammalian autophagy: core molecular machinery and signaling regulation. Current opinion in cell biology, 2010. 22(2): p. 124-131.

175. Yang, Z. and D.J. Klionsky, An Overview of the Molecular Mechanism of Autophagy. Current topics in microbiology and immunology, 2009. 335: p. 1-32.

176. Kabeya, Y., et al., LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. The EMBO journal, 2000. 19(21): p. 5720-8.

177. Kirisako, T., et al., The reversible modification regulates the membrane-binding state of Apg8/Aut7 essential for autophagy and the cytoplasm to vacuole targeting pathway. The Journal of cell biology, 2000. 151(2): p. 263-76.

178. Tanida, I., Autophagy basics. Microbiol Immunol, 2011. 55(1): p. 1-11. 179. Burman, C. and N. Ktistakis, Autophagosome formation in mammalian cells. Seminars in

Immunopathology, 2010. 32(4): p. 397-413.

176

180. Gutierrez, M.G., et al., Rab7 is required for the normal progression of the autophagic pathway in mammalian cells. Journal of cell science, 2004. 117(Pt 13): p. 2687-97.

181. Jager, S., et al., Role for Rab7 in maturation of late autophagic vacuoles. Journal of cell science, 2004. 117(Pt 20): p. 4837-48.

182. Tanida, I., et al., HsAtg4B/HsApg4B/autophagin-1 cleaves the carboxyl termini of three human Atg8 homologues and delipidates microtubule-associated protein light chain 3- and GABAA receptor-associated protein-phospholipid conjugates. The Journal of biological chemistry, 2004. 279(35): p. 36268-76.

183. Pankiv, S., et al., FYCO1 is a Rab7 effector that binds to LC3 and PI3P to mediate microtubule plus end-directed vesicle transport. The Journal of cell biology, 2010. 188(2): p. 253-69.

184. Nixon, R.A., et al., Extensive involvement of autophagy in Alzheimer disease: An immuno-electron microscopy study. Journal of neuropathology and experimental neurology, 2005. 64(2): p. 113-122.

185. Boland, B., et al., Autophagy induction and autophagosome clearance in neurons: relationship to autophagic pathology in Alzheimer's disease. The Journal of neuroscience : the official journal of the Society for Neuroscience, 2008. 28(27): p. 6926-37.

186. Nixon, R.A. and D.-S. Yang, Autophagy failure in Alzheimer's disease—locating the primary defect. Neurobiology of disease, 2011. 43(1): p. 38-45.

187. Hara, T., et al., Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature, 2006. 441(7095): p. 885-9.

188. Pickford, F., et al., The autophagy-related protein beclin 1 shows reduced expression in early Alzheimer disease and regulates amyloid β accumulation in mice. The Journal of clinical investigation, 2008. 118(6): p. 2190-2199.

189. Ravikumar, B., et al., Plasma membrane contributes to the formation of pre-autophagosomal structures. Nature cell biology, 2010. 12(8): p. 747-757.

190. Hailey, D.W., et al., Mitochondria Supply Membranes for Autophagosome Biogenesis during Starvation. Cell, 2010. 141(4): p. 656-667.

191. Axe, E.L., et al., Autophagosome formation from membrane compartments enriched in phosphatidylinositol 3-phosphate and dynamically connected to the endoplasmic reticulum. The Journal of cell biology, 2008. 182(4): p. 685-701.

192. Hamasaki, M., et al., Autophagosomes form at ER-mitochondria contact sites. Nature, 2013. 495(7441): p. 389-93.

193. Berger, Z., et al., Rapamycin alleviates toxicity of different aggregate-prone proteins. Human molecular genetics, 2006. 15(3): p. 433-42.

194. Ravikumar, B., R. Duden, and D.C. Rubinsztein, Aggregate-prone proteins with polyglutamine and polyalanine expansions are degraded by autophagy. Human molecular genetics, 2002. 11(9): p. 1107-17.

195. Spilman, P., et al., Inhibition of mTOR by rapamycin abolishes cognitive deficits and reduces amyloid-beta levels in a mouse model of Alzheimer's disease. PloS one, 2010. 5(4): p. e9979.

196. Falke, L.L., et al., Local therapeutic efficacy with reduced systemic side effects by rapamycin-loaded subcapsular microspheres. Biomaterials, 2015. 42: p. 151-60.

197. Braidy, N., et al., Recent rodent models for Alzheimer's disease: clinical implications and basic research. J Neural Transm, 2012. 119(2): p. 173-95.

198. Kalueff, A.V., A.M. Stewart, and R. Gerlai, Zebrafish as an emerging model for studying complex brain disorders. Trends Pharmacol Sci, 2014. 35(2): p. 63-75.

199. Kumar, S. and S.B. Hedges, A molecular timescale for vertebrate evolution. Nature, 1998. 392(6679): p. 917-20.

177

200. Howe, K., et al., The zebrafish reference genome sequence and its relationship to the human genome. Nature, 2013. 496(7446): p. 498-503.

201. Draper, B.W., P.A. Morcos, and C.B. Kimmel, Inhibition of zebrafish fgf8 pre-mRNA splicing with morpholino oligos: a quantifiable method for gene knockdown. Genesis, 2001. 30(3): p. 154-6.

202. Hwang, W.Y., et al., Efficient genome editing in zebrafish using a CRISPR-Cas system. Nat Biotech, 2013. 31(3): p. 227-229.

203. Huang, P., et al., Heritable gene targeting in zebrafish using customized TALENs. Nat Biotech, 2011. 29(8): p. 699-700.

204. Auer, T.O., et al., Highly efficient CRISPR/Cas9-mediated knock-in in zebrafish by homology-independent DNA repair. Genome Res, 2014. 24(1): p. 142-53.

205. Musa, A., H. Lehrach, and V. Russo, Distinct expression patterns of two zebrafish homologues of the human APP gene during embryonic development. Development genes and evolution, 2001. 211(11): p. 563-567.

206. Leimer, U., et al., Zebrafish (Danio rerio) presenilin promotes aberrant amyloid beta-peptide production and requires a critical aspartate residue for its function in amyloidogenesis. Biochemistry, 1999. 38(41): p. 13602-9.

207. Groth, C., et al., Identification of a second presenilin gene in zebrafish with similarity to the human Alzheimer's disease gene presenilin2. Development genes and evolution, 2002. 212(10): p. 486-90.

208. Campbell, W.A., et al., Zebrafish lacking Alzheimer presenilin enhancer 2 (Pen-2) demonstrate excessive p53-dependent apoptosis and neuronal loss. Journal of neurochemistry, 2006. 96(5): p. 1423-40.

209. Chen, M., R.N. Martins, and M. Lardelli, Complex splicing and neural expression of duplicated tau genes in zebrafish embryos. Journal of Alzheimer's disease : JAD, 2009. 18(2): p. 305-17.

210. Babin, P.J., et al., Both apolipoprotein E and A-I genes are present in a nonmammalian vertebrate and are highly expressed during embryonic development. Proceedings of the National Academy of Sciences of the United States of America, 1997. 94(16): p. 8622-7.

211. Woods, I.G., et al., The zebrafish gene map defines ancestral vertebrate chromosomes. Genome Res, 2005. 15(9): p. 1307-14.

212. Nornes, S., Developmental control of Presenilin1 expression, endoproteolysis, and interaction in zebrafish embryos. Experimental cell research, 2003. 289(1): p. 124-132.

213. van Tijn, P., et al., Presenilin mouse and zebrafish models for dementia: focus on neurogenesis. Progress in neurobiology, 2011. 93(2): p. 149-64.

214. Shen, J., et al., Skeletal and CNS defects in Presenilin-1-deficient mice. Cell, 1997. 89(4): p. 629-39.

215. Herreman, A., et al., Presenilin 2 deficiency causes a mild pulmonary phenotype and no changes in amyloid precursor protein processing but enhances the embryonic lethal phenotype of presenilin 1 deficiency. Proceedings of the National Academy of Sciences of the United States of America, 1999. 96(21): p. 11872-7.

216. Makky, K., J. Tekiela, and A.N. Mayer, Target of rapamycin (TOR) signaling controls epithelial morphogenesis in the vertebrate intestine. Developmental biology, 2007. 303(2): p. 501-13.

217. Tsukamoto, S., et al., Autophagy is essential for preimplantation development of mouse embryos. Science, 2008. 321(5885): p. 117-20.

218. He, C., et al., Assaying autophagic activity in transgenic GFP-Lc3 and GFP-Gabarap zebrafish embryos. Autophagy, 2009. 5(4): p. 520-6.

219. Newman, M., et al., Differential, dominant activation and inhibition of Notch signalling and APP cleavage by truncations of PSEN1 in human disease. Human molecular genetics, 2013.

178

220. Bharadwaj, P.R., et al., Latrepirdine (dimebon) enhances autophagy and reduces intracellular GFP-Abeta42 levels in yeast. Journal of Alzheimer's disease : JAD, 2012. 32(4): p. 949-67.

221. Steiner, H., et al., The biological and pathological function of the presenilin-1 Deltaexon 9 mutation is independent of its defect to undergo proteolytic processing. The Journal of biological chemistry, 1999. 274(12): p. 7615-8.

222. Winkler, E., et al., Generation of Alzheimer Disease-associated Amyloid β(42/43) Peptide by γ-Secretase Can Be Inhibited Directly by Modulation of Membrane Thickness. The Journal of biological chemistry, 2012. 287(25): p. 21326-21334.

223. Lim, A., et al., Analysis of nicastrin gene phylogeny and expression in zebrafish. Development genes and evolution, 2015.

224. Herreman, A., et al., gamma-Secretase activity requires the presenilin-dependent trafficking of nicastrin through the Golgi apparatus but not its complex glycosylation. Journal of cell science, 2003. 116(Pt 6): p. 1127-36.

225. Niimura, M., et al., Aph-1 Contributes to the Stabilization and Trafficking of the γ-Secretase Complex through Mechanisms Involving Intermolecular and Intramolecular Interactions. Journal of Biological Chemistry, 2005. 280(13): p. 12967-12975.

226. Crystal, A.S., et al., Presenilin modulates Pen-2 levels posttranslationally by protecting it from proteasomal degradation. Biochemistry, 2004. 43(12): p. 3555-63.

227. Bedell, V.M., S.E. Westcot, and S.C. Ekker, Lessons from morpholino-based screening in zebrafish. Briefings in Functional Genomics, 2011.

228. Raturi, A. and T. Simmen, Where the endoplasmic reticulum and the mitochondrion tie the knot: the mitochondria-associated membrane (MAM). Biochimica et biophysica acta, 2013. 1833(1): p. 213-24.

229. La Rosa, L.R., et al., Y(682)G Mutation of Amyloid Precursor Protein Promotes Endo-Lysosomal Dysfunction by Disrupting APP–SorLA Interaction. Frontiers in Cellular Neuroscience, 2015. 9: p. 109.

230. Mehmedbasic, A., et al., SorLA complement-type repeat domains protect the amyloid precursor protein against processing. The Journal of biological chemistry, 2015. 290(6): p. 3359-76.

231. Nilsson, P., et al., Abeta secretion and plaque formation depend on autophagy. Cell Rep, 2013. 5(1): p. 61-9.

232. Zhu, X.C., et al., Autophagy modulation for Alzheimer's disease therapy. Mol Neurobiol, 2013. 48(3): p. 702-14.

233. Chau, S., et al., Latrepirdine for Alzheimer's disease. Cochrane Database Syst Rev, 2015. 4: p. CD009524.

234. Bachurin, S., et al., Antihistamine agent Dimebon as a novel neuroprotector and a cognition enhancer. Annals of the New York Academy of Sciences, 2001. 939: p. 425-35.

235. Jones, R.W., Dimebon disappointment. Alzheimer's research & therapy, 2010. 2(5): p. 25. 236. Lermontova, N.N., et al., Dimebon improves learning in animals with experimental Alzheimer's

disease. Bulletin of experimental biology and medicine, 2000. 129(6): p. 544-6. 237. Wu, J., Q. Li, and I. Bezprozvanny, Evaluation of Dimebon in cellular model of Huntington's

disease. Molecular neurodegeneration, 2008. 3: p. 15. 238. Yu, W.H., et al., Macroautophagy—a novel β-amyloid peptide-generating pathway activated

in Alzheimer's disease. The Journal of cell biology, 2005. 171(1): p. 87-98. 239. Tian, Y., et al., Adaptor complex AP2/PICALM, through interaction with LC3, targets

Alzheimer's APP-CTF for terminal degradation via autophagy. Proceedings of the National Academy of Sciences of the United States of America, 2013. 110(42): p. 17071-17076.

179

240. Puglielli, L., R.E. Tanzi, and D.M. Kovacs, Alzheimer's disease: the cholesterol connection. Nature neuroscience, 2003. 6(4): p. 345-51.

241. Huttunen, H.J., C. Greco, and D.M. Kovacs, Knockdown of ACAT-1 reduces amyloidogenic processing of APP. FEBS letters, 2007. 581(8): p. 1688-1692.

242. Puglielli, L., et al., Acyl-coenzyme A: cholesterol acyltransferase modulates the generation of the amyloid [beta]-peptide. Nature cell biology, 2001. 3(10): p. 905-912.

243. Hansson Petersen, C.A., et al., The amyloid beta-peptide is imported into mitochondria via the TOM import machinery and localized to mitochondrial cristae. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(35): p. 13145-50.

244. Picone, P., et al., Mitochondrial dysfunction: different routes to Alzheimer's disease therapy. Oxid Med Cell Longev, 2014. 2014: p. 780179.

245. Chen, J.X. and S.S. Yan, Role of mitochondrial amyloid-beta in Alzheimer's disease. Journal of Alzheimer's disease : JAD, 2010. 20 Suppl 2: p. S569-78.

246. Turrens, J.F., Mitochondrial formation of reactive oxygen species. The Journal of Physiology, 2003. 552(Pt 2): p. 335-344.

247. Gaté, L., et al., Oxidative stress induced in pathologies: The role of antioxidants. Biomedicine & Pharmacotherapy, 1999. 53(4): p. 169-180.

248. Mulligan, V.K. and A. Chakrabartty, Protein misfolding in the late-onset neurodegenerative diseases: common themes and the unique case of amyotrophic lateral sclerosis. Proteins, 2013. 81(8): p. 1285-303.

249. Braakman, I., J. Helenius, and A. Helenius, Role of ATP and disulphide bonds during protein folding in the endoplasmic reticulum. Nature, 1992. 356(6366): p. 260-2.

250. Mirazimi, A. and L. Svensson, ATP is required for correct folding and disulfide bond formation of rotavirus VP7. Journal of virology, 2000. 74(17): p. 8048-52.

251. Contreras, L., et al., Mitochondria: the calcium connection. Biochimica et biophysica acta, 2010. 1797(6-7): p. 607-18.

252. Yi, M., D. Weaver, and G. Hajnoczky, Control of mitochondrial motility and distribution by the calcium signal: a homeostatic circuit. The Journal of cell biology, 2004. 167(4): p. 661-72.

253. Graier, W.F., M. Frieden, and R. Malli, Mitochondria and Ca(2+) signaling: old guests, new functions. Pflugers Archiv : European journal of physiology, 2007. 455(3): p. 375-96.

254. Csordas, G., et al., Structural and functional features and significance of the physical linkage between ER and mitochondria. The Journal of cell biology, 2006. 174(7): p. 915-21.

255. Hedskog, L., et al., Modulation of the endoplasmic reticulum-mitochondria interface in Alzheimer's disease and related models. Proceedings of the National Academy of Sciences of the United States of America, 2013. 110(19): p. 7916-21.

256. Palay, S.L., SYNAPSES IN THE CENTRAL NERVOUS SYSTEM. The Journal of Biophysical and Biochemical Cytology, 1956. 2(4): p. 193-202.

257. Pettegrew, J.W., et al., Alterations of cerebral metabolism in probable Alzheimer's disease: a preliminary study. Neurobiology of aging, 1994. 15(1): p. 117-32.

180

APPENDIX I

Newman, M., Wilson, L., Verdile, G., Lim, A., Khan, I., Moussavi Nik, S.H., Pursglove, S., Chapman, G., Martins, R.N. & Lardelli, M. (2014). Differential, dominant activation and inhibition of Notch signalling and APP cleavage by truncations of PSEN1 in human disease. Human Molecular Genetics, 23(3), 602-617.

NOTE:

This publication is included after page 180 in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1093/hmg/ddt448