Mitochondrial Dynamic Abnormalities in Alzheimer's Disease Sirui Jiang Case Western Reserve University

MITOCHONDRIAL DYNAMIC ABNORMALITIES IN ALZHEIMER’S DISEASE

SIRUI JIANG

Submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Dissertation Advisor: Dr. Xiongwei Zhu

Department of Pathology

CASE WESTERN RESERVE UNIVERSITY

January 2019

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

SIRUI JIANG

Candidate for the degree of Doctor of Philosophy*

Dr. Shu Chen (Committee Chair)

Dr. Xiongwei Zhu

Dr. Xinglong Wang

Dr. George Dubyak

Dr. Charles Hoppel

August 15, 2018 *We also certify that written approval has been obtained for any proprietary material contained therein Table of Contents

Table of Contents 1

List of Figures 3

Acknowledgements 5

List of Abbreviations 7

Abstract 10

Chapter 1. Introduction 12

Introduction to Alzheimer’s Disease 13

General Information 13

Pathology 14

Pathogenesis 15

Introduction to Mitochondrial Dynamics 20

Mitochondrial Function and Neuronal Health 20

Mitochondrial Dynamics 21

Mitochondrial Dynamics and Mitochondrial Function 23

Mitochondrial Dynamics and Mitochondrial Transport 24

Mitochondrial Deficits in AD 26

Mitochondrial Dysfunction in AD 26

Aβ and Mitochondrial Dysfunction 27

Mitochondrial Dynamic Abnormalities in AD: Recent Advances 28

Conclusion 34

Chapter 2. Mfn2 ablation causes an oxidative stress response and eventual

neuronal death in the hippocampus and cortex 36

Abstract 37

Background 39

Methods 43

Results 47

Discussion 54

Figures 60

Chapter 3. DLP1 Cleavage by Calpain in Alzheimer’s Disease 71

Abstract 72

Background 73

Methods 77

Results 80

Discussion 85

Figures 89

Chapter 4. Summary, Discussion and Future Directions 96

References 108

List of Figures

Figure 2.1 Cre-mediated ablation of Mfn2 expression in the hippocampus

and cortex of Mfn2 cKO mice 60

Figure 2.2 Quantification of DLP1 and OPA1 in cKO mice 61

Figure 2.3 Mfn2 ablation caused mitochondrial fragmentation and

ultrastructural damage in the hippocampus in vivo as evidenced by

electron microscopic analysis. 62

Figure 2.4 Mfn2 ablation caused abnormal mitochondrial distribution in vivo. 63

Figure 2.5 Mfn2 ablation caused mitochondrial dysfunction in the brain of

Mfn2 cKO mice. 64

Figure 2.6 Mfn2 ablation caused neurodegeneration in the hippocampus

and cortex in vivo. 65

Figure 2.7 Nissl staining of the cortex 66

Figure 2.8 Mfn2 ablation caused increased oxidative stress in the

hippocampus and cortex in vivo. 67

Figure 2.9 Mfn2 ablation caused increased neuroinflammation in

hippocampus and cortex in vivo. 68

Figure 2.10 Quantification of immunostaining of GFAP, IBA-1, and MAP2 69

Figure 2.11 Mfn2 ablation caused abnormal cytoskeletal alterations in

hippocampus and cortex in vivo. 70

Figure 3.1 Dose- and time-dependent cleavage of recombinant DLP1

by calpain-1. 89

Figure 3.2 DLP1 is cleaved by calpain in M17 neuroblastoma cell lysates after

incubation with calpain-1. 90

Figure 3.3 Calpain-dependent cleavage of spectrin and DLP1 in

glutamate-treated rat primary cortical neurons. 91

Figure 3.4 Calpain-dependent cleavage of spectrin and DLP1 in rat

primary cortical neurons treated with soluble Aβ oligomers. 92

Figure 3.5 Calpain-dependent cleavage of spectrin and DLP1 in rat

primary cortical neurons treated with okadaic acid. 93

Figure 3.6 Calpain-dependent cleavage of spectrin and DLP1 in

primary cortical neurons isolated from CRND8 APP transgenic mice. 94

Figure 3.7 Decreased level of DLP1 in Alzheimer’s Disease (AD) brain. 95

Acknowledgements

First and foremost, I would sincerely like to thank my thesis advisor, Dr. Xiongwei Zhu, for his never ending support throughout my training in this doctoral program. Dr. Zhu has always been a source of great assistance and advice whether it is science or my future career. He has not only taught me how to think as a scientist but also encouraged me to approach scientific questions with a well thought out and critical methodology. This thesis would not be possible without his extensive expertise and immense amount of patience and support.

I would also like to thank my co-mentor, Dr. Xinglong Wang, for his impeccable advice and guidance through my training. He is always approachable for any technical or methodological questions in regards to my research. He has also taught me how to approach a scientific problem from many angles and develop the proper methods in order to tackle any kind of problem. Of course this work could not have been completed without the help and assistance of my laboratory colleagues such as Dr. Wenzhang

Wang, Dr. Xiaopin Ma, Sandra Siedlak, Sandy Torres, Priya Nandy, and many others who have contributed to my learning and research.

I would like to thank my thesis committee: Dr. Robert Petersen, Dr. Shu Chen, Dr.

George Dubyak, Dr. Xinglong Wang, and Dr. Charles Hoppel for their continued guidance and advice through my doctoral training.

I would like to acknowledge the financial support of the National Institute of Health grant

T32 GM007250, T32 NS077888, R01 NS083385, Dr. Robert M. Kohrman Memorial

Fund, and the Department of Pathology at Case Western Reserve University.

Finally, I would like to thank my parents and sister who have always been a source of unconditional love, comfort, and support for me throughout this long journey. They have always inspired me to pursue my dreams and have been with me every step of the way and for that, I am forever grateful.

List of Abbreviations

AD Alzheimer's disease

ADDLs amyloid-beta derived diffusible ligands

ApoE Apolipoprotein E

APP Amyloid precursor protein

Aβ Amyloid-β

C83 C-terminal fragment of 83 amino acids

C99 C-terminal fragment of 99 amino acids

CaMKII-Cre+/-/Mfn2loxP/loxP Mfn2 conditional knockout mouse

Cdk5 cyclin-dependent protein kinase 5

CHOP C/EBP homologous protein

cKO Conditional knockout

CNS Central nervous system

CREB cAMP response element-binding protein

DIV Days in vitro

DLP1 Dynamin-like protein

DNP 2,4-Dinitrophenol

EM Electron microscopy

ER Endoplasmic Reticulum fAD Familial AD

Fis1 Mitochondrial fission 1 protein

GDAP1 Ganglioside induced differentiation associated protein 1

GFAP Glial fibrillary acidic protein

GSK-3β Glycogen-synthase kinase-3β

GWAS Genome wide association studies

HBSS HEPES-buffered salt solution

IBA-1 Ionized calcium-binding adapter molecule 1

LOAD Late-onset AD

LTP Long-term potentiation

MAM Mitochondria associated membrane

MAP1/2 Microtubule associated proteins ½

MAP2 Microtubule-associated protein-2

MAPT Microtubule associated protein tau

Mff Mitochondrial fission factor

Mfn1 Mitofusin 1

Mfn2 Mitofusin 2

Mid49 Mitochondrial elongation factor 2

Mid51 Mitochondrial elongation factor 1 mtDNA Mitochondrial DNA

NeuN Neuronal nuclei

NFTs Neurofibrillary tangles

NO Nitric oxide

OCR Oxygen consumption rate

OPA1 Optic atrophy protein 1

OXPHOS Oxidative phosphorylation

PBS Phosphate-buffered saline

PET Positron emission tomography

PHFs Paired helical filaments

PKA Protein kinase A

PP-2A Protein phosphatase-2A

PS1 Presenilin 1

PS2 Presenilin 2

ROS Reactive oxygen species

RyR Ryanodine receptor

sAD Sporadic AD

SNPs Single nucleotide polymorphisms

SP Senile plaques

TPR Tetratricopeptide

TUNEL Terminal deoxynucleotidyl transferase nick-end labeling

Mitochondrial Dynamic Abnormalities in Alzheimer’s Disease

Abstract

SIRUI JIANG

Alzheimer’s Disease (AD) is the most common cause of dementia leading to progressive memory loss and neurodegeneration in the hippocampus and frontal cortex. While it has been shown that mitochondrial dysfunction is an early and prominent feature in the progression of AD, it is unclear whether mitochondrial dysfunction itself can lead to neurodegeneration in AD-affected brain regions. Evidence has already suggested that various mitochondrial dynamic proteins (DLP1, OPA1, Mfn1, Mfn2, Fis1) are altered in

AD and that there is an imbalance of mitochondrial fission and fusion yet there is a knowledge gap of whether this altered dynamics leads to neurodegeneration and what mechanisms lead to altered mitochondrial proteins in AD. To answer these questions, we created an Mfn2 conditional knockout mouse to recapitulate mitochondrial fragmentation phenotype in the hippocampus and frontal cortex. We found that indeed loss of mitochondrial fusion leads to mitochondrial morphological and bioenergetics abnormalities. These early changes lead to a series of events including oxidative stress, inflammation, and microtubule abnormalities that precede neurodegeneration. To understand the underlying mechanisms leading to loss of important mitochondrial dynamic proteins in AD, we treated primary neurons with amyloid-beta derived diffusible ligands to mimic AD and we found that the loss of DLP1 and Mfn2 is attributed to the calcium-activated protease, calpain. We also found that calpain specifically cleaves DLP1 leading the appearance of several cleavage fragments in both AD transgenic mice as well

10 as AD patient brains. Altogether, these studies show that loss of mitochondrial dynamics could lead to neurodegeneration and that the loss of mitochondrial dynamic proteins in

AD could be through the activity of calcium-activated proteases such as calpain.

Chapter 1. Introduction

Alzheimer’s Disease

General Information

Alzheimer’s Disease (AD) is the most common neurodegenerative disorder of the

elderly that leads to progressive memory loss, impairments in behavior and language, and

ultimately death. As the most prevalent form of dementia, AD affects nearly 5.7 million

Americans and over 50 million people worldwide (Alzheimer’s Association, 2018). This

disease was first observed and reported in 1906 by Dr. Alois Alzheimer, a German

physician for whom the disease is named after, who described the symptoms of a 51-year

old woman named Auguste D as she suffered from memory disturbances, paranoia, and

progressive confusion (Reck & Mach 1988). In his report Dr. Alzheimer noted distinct

plaques and tangles in the brain histology. The major brain areas affected in this disease

are: cerebral cortex controlling consciousness and ideas, hippocampus controlling new

memory formation, and the basal forebrain controlling memory and learning (Alloul et al.

1998). Clinically, AD is diagnosed by examination of patients showing memory loss,

inability to tell time and place, language impairment, and a progressive cognitive

decline(Förstl & Kurz 1999). The prevalence of AD in people >70 years of age in the US

is 9.7% (Plassman et al. 2007), while worldwide the global prevalence of dementia in

people >60 years of age is 3.9%. These staggering numbers will only increase with

around 5 million new cases every year (Brookmeyer et al. 2007). The risk of AD tends to

increase with age and patients are usually divided into two groups: early-onset AD for patients < 65 years of age (which make up about 10% of all AD population) and late – onset AD for patients >65 years of age (which make up about 90% of AD population)

(Alzheimer’s Association, 2018). Sporadic AD (sAD; >65 years of age) makes up the

majority of the patient population where the cause of disease is currently unknown.

Whereas, familial AD (fAD; <65 years of age), makes up a small percentage of the AD

population due to specific mutations in important AD-related genes, which gives us an idea of where to start researching on the patho-mechanisms of disease (Alzheimer’s

Association, 2018). The average survival time after diagnosis is 8-10 years (Larson et al.

2004). Currently there is no cure for this disease due to the complexity of disease pathogenesis as well as an unclear mechanism of neurodegeneration (Alzheimer’s

Association, 2018). Although there are FDA-approved drugs (memantine, cholinesterase inhibitors) and other treatments (vitamin E, Omega-3 fatty acids, antioxidants, etc.) that are used to manage the cognitive symptoms in AD patients, none of them have been shown to reverse the pathology of AD (Alzheimer’s Association 2018).

Pathology

As first described and observed by Dr. Alzheimer, AD is historically characterized by the presence of two pathological hallmarks: senile plaques (SPs) composed of amyloid beta and neurofibrillary tangles (NFTs) composed of hyperphosphorylated tau (Smith 1998). Other pathological changes include the presence of neuronal loss, atrophy of the cortical and medial temporal regions of the brain, neuropil threads, dystrophic neurites, granulovacuolar degeneration, and Hirano bodies

(Smith 1998). SPs are spherical extracellular lesions 10-200 μm in diameter with a core made up of 6-10 nm bundles of amyloid-β (Aβ) peptides (Smith 1998).Whereas NFTs are intracellular aggregates of paired helical filaments (PHFs) composed of hyperphosphorylated tau. The number of NFTs are highly correlated with the degree of

dementia which suggests that these NFTs correlate with neuronal dysfunction and

degeneration (Brion 1998). Braak staging has been used to define the progression of

NFTs and is used both in research as well as clinical diagnosis of AD due to their

relatively consistent distribution pattern amongst AD patients as opposed to the widely

varied SP distribution. Stage I-II are when the NFTs involvement is confined mainly in

the transentorhinal region, stage III-IV involve the limbic regions, stage V-VI are

extensively neocortical involvement (Braak & Braak 1991).

Pathogenesis

Even though the end stage pathologic hallmarks of AD are well documented and

studied, the exact mechanism or mechanisms leading to the formation of SPs and NFTs

in the brain or neurodegeneration has not been identified. Although many risk factors

have been detailed and explored such as aging, SP/NFT formation, inflammation,

oxidative stress, mitochondrial dysfunction, and cell cycle abnormalities (Swerdlow

2007), the number one risk factor for development of AD is age, as > 90% of AD patients

are 65 years or older with the likelihood of developing AD nearly doubling every 5 years after the age of 65 and reaching nearly 50% at age 85 (Alzheimer’s Association, 2018).

This strong correlation between age and AD development seems to suggest that factors both systemically and environmentally lead to the pathogenesis of sporadic AD. Through genome wide association studies (GWAS), it has been established that inflammatory and immune responses may play a role in the development and progression of AD (Jones et al. 2015). It has also been suggested that the accumulation of free radicals or oxidative stress over time lead to damage to DNA, proteins, lipids, and sugars and contribute to the

pathogenesis of AD (Nunomura et al. 2006). Genetics studies have also linked

apolipoprotein E (ApoE) as a high risk factor for development of late-onset AD (Pericak-

Vance et al. 1991). The exact mechanism of how aging leads to an increased risk for

development of AD has yet to be elucidated. However, through genetic studies of

mutations found in fAD cases point us in the direction of amyloid beta.

In familial AD cases, which encompass < 10% of all AD cases, a mutation is

involved in at least one of three transmembrane proteins: amyloid precursor protein

(APP), presenilin 1 (PS1), or presenilin 2 (PS2) (Tang & Gershon 2003). The APP gene

is located on chromosome 21 , and intriguingly patients with Down syndrome or Trisomy

21 (triplicate copies of chromosome 21) develop AD pathology early (Prasher et al.

1998), which lead us to believe that APP dosage is involved in AD pathogenesis. To date,

there have been at least 24 mutations found in the APP gene that lead to development of

AD (De Jonghe 2001). Most of the mutations in APP are located near the cleavage sites

of the secretase enzymes (König et al. 1992; De Jonghe 2001) which suggests that the

abnormal proteolysis of APP is what leads to AD. PS1 and PS2 are two highly

homologous proteins that are crucial for γ-secretase function with more than 180 mutations in PS1 and more than 10 mutations in PS2 that have been found to cause AD

(Bagyinszky et al. 2014). Mutations in APP or PS1/2 lead to either increased production

of total Aβ or increased Aβ1-42/Aβ1-40 ratio. These findings led to the “amyloid cascade

hypothesis” suggesting that the accumulation of amyloid β through the mutations

mentioned above is what ultimately leads to AD (Hardy & Higgins 1992).

APP is a type I integral membrane protein with the Aβ domain partially

embedded in the membrane. APP protein is processed by several different secretase

proteins occurring in two distinct pathways: the non- amyloidogenic and amyloidogenic

pathways. In the non-amyloidogenic pathway, full length APP is first cleaved by α-

secretase within the Aβ domain which releases the soluble α-APP into the extracellular matrix, and generates a membrane-bound C-terminal fragment of 83 amino acids (C83)

(Newman et al. 2007). C83 is then cleaved by γ-secretase in the middle of the transmembrane domain to produce a 3-KD peptide called p3. In the amyloidogenic pathway, APP is first cleaved by β-secretase at the N-terminus of Aβ domain, releasing the soluble β-APP in to the extracellular matrix and generates a membrane-bound 99- amino acid C-terminal fragments (C99). C99 is then cleaved by γ-secretase to generate

Aβ. This cleavage results in two different length Aβ peptides, 1-40 and 1-42, the former being the major product and the latter created in small quantities (Newman et al. 2007).

The longer Aβ1-42 is more hydrophobic and much more prone to fibril formation than

Aβ1-40 and represents the major Aβ species in SP. Recent studies suggest that the mixture of both Aβ1-40 and Aβ1-42 and their ratio plays an important role in their toxicity and oligomerization on neurites (Sengupta et al. 2016). It is widely believed that the accumulation of soluble Aβ1-42 oligomers induces neuronal cytotoxicity leading to neurodegeneration in the brain (Newman et al. 2007; Kirkitadze & Kowalska 2005;

Dahlgren et al. 2002). One possible way that has been suggested that Aβ peptides produce toxicity is by membrane disruption via pore formation and ion dysregulation

(Kayed & Lasagna-Reeves 2012). Another mechanism is through Aβ oligomers directly interacting with receptors such as inhibiting long-term potentiation (LTP) response

through N-methyl-D-aspartate (NMDA) receptors (Li et al. 2011), binding to prion

protein (PrP) and disrupting synaptic function (Freir et al. 2011; Laurén et al. 2009; Um

et al. 2012), or through rapid depolarization of neurons leading to glutamate release and

excitotoxicity due to calcium influx (Morkuniene et al. 2015).

There are currently no mutations in microtubule associated protein tau (MAPT)

associated with AD, however abnormal aggregation of the hyperphosphorylated tau is

central to AD neurofibrillary pathology. In the normal human brain there are six

molecular isoforms of tau (Goedert et al. 1989) and these isoforms interact with tubulin

to promote its assembly into microtubules necessary for stabilizing neuronal structure

(Weingarten et al. 1975). Like the other microtubule associated proteins (MAP1/MAP2),

tau is regulated by its degree of phosphorylation (Lindwall & Cole 1984; Alonso et al.

1994). Normal brain tau contains 2-3 moles of phosphate per mole of protein (Schober et

al. 2012), whereas the abnormal hyperphosphorylated tau has >6 moles of phosphate per

mole of protein (Iqbal et al. 2010). Normally tau can be hyperphosphorylated to down

regulate its activity during development as well as during hibernation and hypothermia

(Schmitt et al. 1977; Su et al. 2008), yet in AD such control of appropriate level of tau

phosphorylation/dephosphorylation by regulation of the proper balance between kinase

and phosphatase appears lost. It is believed that the activity of protein phosphatase-2A

(PP-2A) is responsible for the regulation of tau de-phosphorylation and its decreased

activity leads to the accumulation of hyperphosphorylated tau and the downstream

neuropathology (Liang et al. 2008). Increased activity of tau kinases have also been implicated in the dysregulation of tau phosphorylation in AD which includes glycogen- synthase kinase-3β (GSK-3β), cyclin-dependent protein kinase 5 (cdk5), cAMP-

dependent protein kinase (PKA), and stress-activated protein kinases (Ferrer et al. 2005;

Mazanetz & Fischer 2007). Hyperphosphorylated tau sequesters normal MAPs and disrupts microtubules and it also self-assembles into paired helical filaments (PHFs) which make up the bulk of NFTs (Alonso et al. 1994; Iqbal et al. 2010). Recent work in

animal models has shown that tau is required for Aβ toxicity in vivo where knockout of tau confers protection from learning and memory deficits in APP mutant mice (Roberson et al. 2007).

Apolipoprotein E (ApoE) has been found to increase the risk of developing late- onset AD (LOAD) (Corder et al. 1993). ApoE is a 34 kD glycoprotein that functions as a ligand in the endocytosis of lipoproteins (Kim et al. 2009). It is most abundant in the liver but can also be expressed in non-neuronal cells like astrocytes and microglia (Mahley et al. 2009). It is the single-nucleotide polymorphisms (SNPs) that produces three common but slightly different ApoE isoforms: ApoE2, ApoE3, ApoE4 (Mahley et al. 2006).

Possession of the ApoE4 allele is a strong risk factor for the development of AD with a 2-

3 fold increase for those possessing one allele and a 12 fold increase in those possessing two alleles (Kim et al. 2009). ApoE has been shown to play a role in accumulation and deposition of Aβ as levels of Aβ and plaque loads are ApoE isoform-dependent (E4 >

E3 > E2) (Reiman et al. 2009; Castellano et al. 2011; Bales et al. 2009). However, the exact mechanisms by which ApoE isoforms regulate Aβ aggregation and deposition require further investigation.

Mitochondrial Dynamics

Mitochondrial Function and Neuronal Health

Mitochondria are double membrane organelles found in most eukaryotic cells.

These small organelle are essential to many cellular functions, because they produce cellular ATP, synthesize key cellular metabolites, regulate apoptosis, buffer calcium, and produce endogenous reactive oxygen species (ROS) (Frey & Mannella 2000; Benard et al. 2007). Besides their important cellular functions, mitochondria are very dynamic organelles, which can be rapidly transported to compartments of physiological need and they undergo constant division or fusion (Chan 2006). Due to their high metabolic demand, neurons are particularly dependent on mitochondria to provide the ATP required for normal function (Kann & Kovács 2007). Mitochondria also help neurons in maintaining calcium homeostasis which is critical for proper synaptic function (Mattson et al. 2008; MacAskill et al. 2009) through the shuttling of mitochondria to areas of large ion flux which is maintained by mitochondrial transport proteins. Mitochondria are involved in ATP supply to cells through oxidative phosphorylation (OXPHOS) which contains many redox enzymes and naturally occurring inefficiencies of oxidative phosphorylation generate reactive oxygen species (ROS). Mutations in mitochondrial

DNA (mtDNA), generation and presence of ROS, and environmental factors may contribute to energy failure and lead to neurodegenerative diseases (Federico et al. 2012;

Alexander et al. 2000; Züchner et al. 2004).

Mitochondrial Dynamics

The term mitochondrial dynamics refers to how mitochondria can quickly change

their morphology, length, size, number, and location within a cell based on responses to

cellular necessities. Mitochondrial dynamics is controlled by a delicate balance of two opposing forces in mitochondrial fission and fusion (Chan 2006). Frequent fusion and fission events serve as an efficient means of maintaining mitochondrial number and morphology (Fox 2012). Inhibition of mitochondrial fusion leads to appearance of small

“fragmented” mitochondria whereas inhibition of fission results in elongated mitochondria (Bleazard et al. 1999; Sesaki & Jensen 1999). Mitochondrial fission involves several proteins: a large GTPase, dynamin-like protein 1 protein (DLP1, also referred to as DRP1 or DNM1L), mitochondrial fission 1 protein (Fis1), mitochondrial fission factor (Mff), mitochondrial elongation factor 2 (Mid49 or MIEF2), mitochondrial elongation factor 1 (Mid51 or Mief1), and ganglioside induced differentiation associated protein 1 (GDAP1) (Knott et al. 2008; van der Bliek et al. 2013). The majority of DLP1 resides in the cytoplasm and during fission it is recruited to the mitochondrial surface via proteins such as Fis1, Mff, Mid49, and Mid51/Mief. At the sites of fission, it has been shown that DLP1 forms a unique helical assembly that constricts mitochondrial membranes much like the dynamin protein (Mears et al. 2011; Smirnova et al. 2001).

Fis1 resides on the surface of mitochondria where its N-terminus forms a tetratricopeptide (TPR)-like domain and recruits DLP1 to the mitochondrial surface

(James et al. 2003; Wells et al. 2007). Mff is a relatively small protein and similar to

Fis1, it has a transmembrane C-terminus that anchors the protein in the mitochondrial outer membrane, acts a DLP1 receptor much like Fis1 (Otera et al. 2016). Recent work

has shown that Fis1, Mff, Mid 49, and Mid51/Mief all act in a partially redundant way to

promote DLP1 dependent fission (Losón et al. 2013) although further studies must be

done to elucidate their mechanisms. The end product of fission are two daughter

mitochondria that have remodeled cristae and other structures.

In opposition to fission is mitochondrial fusion which is regulated by three large

GTPase proteins: Mitofusin 1 (Mfn1, also referred to as hfzo1), Mitofusin 2 (Mfn2, also

referred to as CMT2A2, CPRP1 and MARF) and optic atrophy protein 1 (OPA1, also

referred to as MGM1, NPG and NTG) (Chan 2006; Knott et al. 2008). Mfn1 and Mfn2

are mitochondrial transmembrane proteins localized to the outer mitochondrial membrane

where they function independently to promote outer membrane fusion (Chen et al. 2005).

OPA1 is localized to the inner mitochondrial membrane and is thought to be responsible

for inner membrane fusion. Fusion involves the tethering of two adjacent mitochondria

through a dimeric antiparallel coiled-coil structure forming either homotypic (Mfn1-

Mfn1 or Mfn2-Mfn2) or heterotypic (Mfn1-Mfn2) dimers (Koshiba 2004; Ishihara et al.

2004). This type of molecular symmetry required for outer membrane fusion is not required for inner membrane fusion. OPA1 exists within the mitochondrial inner membrane space as both a membrane-bound long form (L-OPA1) as well as a soluble short form (S-OPA1) and inner membrane fusion requires a combination of both L-OPA1 and S-OPA1 (Song et al. 2007). Fusion of the inner membrane mediated by OPA1 also requires the presence of cardiolipin presenting an interaction that differs from the homotypic or heterotypic interactions of outer membrane fusion (Ban et al. 2017).

The regulation of mitochondrial dynamics is important in order to respond to various stimuli and demands across the body. Mitochondrial fission is regulated through

post-translational modifications of DLP1 through ubiquitination, sumoylation, and

phosphorylation. Phosphorylation of DLP1 at Ser637 by Protein kinase A (PKA)

deactivates it and inhibits fission, whereas phosphorylation of DLP1 at Ser585 by cyclin-

dependent kinase 1 (CDK1) increases fission (Chang & Blackstone 2007; Cribbs &

Strack 2007; Taguchi et al. 2007). Ubiquitination of DLP1 by mitochondrial ubiquitin

ligase (MITOL) leads to its degradation and consequent decrease in fission (Yonashiro et al. 2006). Sumoylation of DLP1 leads to protection from degradation and an increased translocation of DLP1 to the mitochondrial outer membrane (Harder et al. 2004). Mfn2

activity is regulated by a transcriptional modulator (Smad2) and a guanine nucleotide

exchange factor Rab and Ras Interactor (RIN1). Smad2 recruits RIN1 to Mfn2 where this

complex then allows for the activation of the GTPase function of Mfn2, which then

promotes fusion (Kumar et al. 2016). Opposite to popular belief that Mfn2 is

constituently active, a recent report has shown that Mfn2 exists in two functionally

distinct conformations, a compressed (inactive) and extended (active) form, directed by specific intramolecular interactions (Franco et al. 2016; Chandhok et al. 2018). The regulators of fusion and fission are only beginning to be elucidated and more investigation is necessary to clarify these complex interactions.

Mitochondrial Dynamics and Mitochondrial Function

One question that arises is why must mitochondria expend so much energy

(DLP1, Mfn2/1, OPA1 are large GTPases) in order to fuse or break apart? Mitochondrial fission and fusion events are necessary to maintain a healthy population of mitochondria.

Fusion promotes the exchange of lipid membrane and intra-mitochondrial materials such as mtDNA. Fission promotes sequestration and elimination of irreversibly damaged

mitochondria keeping the integrity of mitochondria intact (Chen et al. 2007; F. Liu et al.

2005; Twig et al. 2008). Fusion and fission leads to dramatic changes in mitochondrial

morphology, but there is no evidence of a direct link between changes in morphology and

mitochondrial dysfunction, although there is growing evidence that suggests keeping the

balance between fission and fusion is important to maintaining mitochondrial functions

including energy production, calcium signaling, ROS production, and apoptosis.

Unbalanced mitochondrial dynamics, either too much fission or too much fusion, results

in decreased energy production and cellular metabolism. Cells deficient in Mfn2/1 or

OPA1 show evidence of mitochondrial dysfunction with fragmented mitochondria and

decreased oxidative respiration rates (Chen et al. 2005; Chen et al. 2003). This finding

was corroborated by fibroblasts from patients with mutations in either Mfn2 or OPA1

showing similar decreases in oxidative phosphorylation (OXPHOS) rates and ATP

production (Casasnovas et al. 2010; Loiseau et al. 2007; Nochez et al. 2009). DLP1

inhibition also showed similar decreases in ATP production and also an elongation of

mitochondria due to impairment of fission (Benard et al. 2007). Mitochondrial fission

(fragmentation) is associated with cell death and is an early event during apoptosis that

precedes cytochrome c release and caspase activation (Frank et al. 2001). Evidence

suggests that fragmented mitochondria do not immediately cause cell death, but rather

sensitizes the cell to apoptotic cell death through mechanisms involving the transient

opening of the mitochondrial permeability transition pore (mPTP) (Picard et al. 2013). In

general, inhibition of fission reduces apoptosis while inhibition of fusion facilitates cell

death (Sugioka et al. 2004; Olichon et al. 2003). Changes in mitochondrial shape not only affects mPTP opening, but also its handling of calcium and mitochondrial calcium

buffering (Szabadkai et al. 2006). Fragmentation of mitochondrial networks lead to significantly greater increase in matrix calcium levels than elongated mitochondria

(Paltauf-Doburzynska et al. 2004).

Mitochondrial Dynamics and Mitochondrial Transport

Mitochondrial dynamics does not just only include fusion and fission, as mitochondria are also highly mobile organelles that travel along the cytoskeleton to position themselves based on areas of high energetic demand or high ionic activity. It has been well documented that changes in the mitochondrial fission/fusion proteins lead to a change in mitochondrial distribution (Smirnova et al. 1998; Spinazzi et al. 2008). For

example, perturbation of DLP1 that leads to mitochondrial elongation or OPA1 that leads

to mitochondrial fragmentation could affect the number of dendritic mitochondria in

primary neurons (Li et al. 2004). Similarly, DLP1 mutations in Drosophila also resulted

in failure to populate the distal axon with mitochondria in vivo (Verstreken et al. 2005).

In fact, Mfn2 has been shown to directly regulate mitochondrial transport as an

interacting protein with the mitochondrial transport proteins Miro and Milton (Misko et

al. 2010). This loss of mitochondrial transport along with loss of fusion (resulting in

fragmented mitochondria) resulted in segmental axonal degeneration, loss of calcium

homeostasis, and accumulation of ROS (Misko et al. 2012). The Miro protein has been

shown to act as a cytosolic calcium sensor through its EF-hand domain 1 and also a

determinant for mitochondrial shape change that is distinct from fusion or fission in

response to cytosolic calcium (Nemani et al. 2018). Further mechanisms involved in the

shaping of mitochondria and how that affects their distribution have yet to be elucidated.

Mitochondrial Deficits in AD

Mitochondrial Dysfunction in AD

Numerous studies have implicated a metabolic defect in the brain as an early and prominent marker in AD (Blass 2000). Positron emission tomography (PET) scans consistently demonstrate reduced cerebral metabolism in the temporoparietal cortices in

AD (Minoshima et al. 1997) and these scans can even show reduced metabolism in AD- related brain regions that precede any evidence of functional impairment by neurological testing or atrophy via imaging (Blass 2000). Due to the fact that mitochondria are almost exclusively the major contributors of brain metabolism, these findings suggest that mitochondrial dysfunction plays an early and important role in the pathogenesis of AD.

There are no clear mitochondrial mutations that lead to AD, yet damaged mitochondria have a decreased ATP output and produce more ROS, both of which are featured in AD (Castellani et al. 2002; Gibson et al. 1998). Components in the mitochondrial metabolism have been consistently shown to be deficient in AD: α- ketoglutarate dehydrogenase complex (KGDHC), pyruvate dehydrogenase complex

(PDHC), and cytochrome oxidase (COX) (Gibson et al. 1998; Cottrell et al. 2001; Maurer

2000; Nagy et al. 1999). Using cybrids, where mitochondria derived from AD patients are fused with mitochondria-depleted cells, metabolic abnormalities, elevated ROS levels and defects in calcium handling have all been observed (Khan et al. 2000; Abramov

2004). Along those lines, spontaneous alterations in mitochondrial DNA (mtDNA) has been shown to be significantly increased in AD (Hirai et al. 2001; Corral-Debrinski et al.

1994). Genetic alterations in mtDNA have been linked to an increased incidence of AD

(Coskun et al. 2004). As sequencing technology has improved over the years, more recent studies have shown that mtDNA mutations increase in early stage AD by using a highly accurate next-generation sequencing methodology (Hoekstra et al. 2016) and these mutations are not the result of oxidative damage but rather replication errors. This contradicts many early studies suggesting that because mitochondria are the major site of

ROS production, mtDNA is a primary target of oxidative damage (Sullivan & Brown

2005). As mtDNA damage accumulates, there is loss of normal mitochondria function and an increased production of ROS due to malfunction in the respiratory chain (Migliore et al. 2005).

Aβ and Mitochondrial Dysfunction

Although cleavage of APP leads to mostly extracellular Aβ, there has been growing animal and human evidence to suggest that this peptide can accumulate intracellularly and contribute to disease progression (LaFerla et al. 2007). Transgenic mouse studies confirm these findings of intracellular Aβ accumulation that precedes the formation of extracellular plaques (Chui et al. 1999; Knobloch et al. 2007; Li et al. 2002).

In neurons, Aβ can be generated at intracellular sites such as the ER (Cook et al. 1997;

Wild-Bode et al. 1997), Golgi (Hartmann et al. 1997), and the endosomal-lysosomal systems (LaFerla et al. 2007). Aβ has been demonstrated in the inner membrane or matrix of mitochondria in neuroblastoma cells overexpressing mutant APP as well as in the postmortem AD brain (Anandatheerthavarada et al. 2003; Caspersen et al. 2005; Manczak et al. 2006).

Studies in APP transgenic mice suggest that mitochondria associated Aβ accumulation leads to impairment of respiratory chain related complexes and alters the expression profile of mitochondrial and apoptotic genes (Reddy et al. 2004). In neurons, it is interesting that the toxicity of the Aβ peptide requires functional mitochondria

(Cardoso et al. 2001). And it has been reported that Aβ oligomers could impair mitochondrial function through elevation of ROS and many mitochondrial respiratory enzymes are vulnerable to ROS (Casley et al. 2001). As noted above, mitochondria morphology changes lead to opening of mPTP which could activate the intrinsic apoptotic pathway or at the very least sensitize neurons to apoptosis. It has been shown that Aβ causes apoptosis through inhibition of complex I of the respiratory chain (Casley et al. 2001). Other studies have implicated a role of phospholipase A2 (PLA2) as a key player in Aβ-induced mitochondrial dysfunction through the NADPH oxidase and mitogen-activated protein kinase pathways (Abramov 2004; Zhu et al. 2006).

Mitochondrial Aβ has been shown to interact with Cyclophilin D (CypD), which is part of the mPTP and promotes opening of this pore leading to apoptosis, lowered mitochondrial calcium uptake, and impaired mitochondrial respiratory function (Du et al.

2011). Mitochondrial transport has been shown to be affected by Aβ as studies have implicated the Glycogen Synthase Kinase 3β (GSK3β) activity as the major culprit in impairing mitochondrial axonal transport (Rui et al. 2006; Decker et al. 2010).

Mitochondrial Dynamic Abnormalities in AD: Recent Advances

As discussed above, the number and size of mitochondria are strictly regulated by mitochondrial dynamics. The first evidence implicating mitochondrial dynamic

abnormalities in AD came from an earlier ultrastructural study on mitochondria in

pyramidal neurons of AD biopsied brain tissues from Perry’s group: in addition to

significant structural damage to mitochondria as evidenced by partial or even total loss of

cristae, AD neurons demonstrated significantly reduced number as well as significantly

increased size of mitochondria compared to age-matched controls (Hirai et al. 2001).

Intrigued by this earlier observation, our group set out to study mitochondrial dynamics

in AD brain and found significant changes in the expression of mitochondrial dynamic

proteins in the hippocampal tissues from AD patients where all the large GTPases

including the fission factor DLP1 and fusion factors Mfn1, Mfn2 and OPA1 are

significantly reduced while the fission factor Fis1 is increased (Wang et al. 2009). It is of

interest to note that these changes, i.e., either reduced DLP1, Mfn1, Mfn2, OPA1 or

increased Fis1, consistently leads to reduced mitochondrial densities in neurites along

with reduced dendritic spines. This is consistent with the finding of an abnormal

mitochondrial distribution pattern in the AD pyamidal neurons where mitochondria are

relatively depleted from neuronal processes. Consistently, later studies from other groups

also found significant alterations in the expression of these mitochondrial dynamic

proteins in AD brains although there are controversies regarding how DLP1 level is

changed: Bossy-Wetzel’s group confirmed reduced DLP1 in AD (Bossy et al. 2010)

while Reddy’s group demonstrated significantly increased DLP1 levels in AD tissues

(Reddy et al. 2012). Nevertheless, it was agreed that there is increased mitochondrial

translocation of DLP1, which likely caused excessive mitochondrial fission that resulted

in mitochondrial fragmentation in AD neurons. Interactions between DLP1 and Aβ or phosphorylated tau in AD hippocampal tissues correlated with the severity of the disease

(Manczak & Reddy 2012). Importantly, more detailed ultrastructural analysis of biopsied

AD tissues revealed significant reduction in mitochondrial length in AD neurons,

confirming a likely involvement of mitochondrial fragmentation in the course of AD

(Wang et al. 2008).

Mitochondrial fragmentation and abnormal distribution and their essential role in

mitochondrial dysfunction and neuronal vulnerability is more unequivocally

demonstrated in APP- or Aβ-related in vitro models of AD by multiple groups:

mitochondria became fragmented and accumulated in the perinuclear area in M17 human

neuroblastoma cells overexpressing wild type APP which was further exacerbated by the

expression of fAD-causing APPswe mutant (Wang et al. 2008). Changes in the

expression of mitochondrial fission/fusion proteins, consistent with a mitochondrial

fragmentation phenotype, were found in these cells. Expression of WT or mutant APP

also caused mitochondrial fragmentation and reduced distribution of mitochondria in

neuronal process in primary neurons which could be rescued by BACE-1 inhibitor.

Exposure to soluble Aβ oligomers also caused a time- and dose-dependent change in the expression of fission/fusion proteins and significant mitochondrial fragmentation and abnormal mitochondrial distribution in neurites. Consistent with such morphological changes, DLP1 translocation to mitochondria is significantly increased in oligomer Aβ -

treated cells that is likely due to changes in the posttranslational regulations of DLP1. In

an early study, it was observed that nitric oxide (NO) treatment in primary neurons lead

to fragmentation of the normally long filamentous mitochondria along with ultrastructural

damage, increased ROS, decreased respiratory rate, and autophagy (Barsoum et al. 2006).

Following this finding, it was observed that NO leads to activation of DLP1 by S-

nitrosylation and that same effect can be recapitulated by treated with an Aβ25-35

fragment or Aβ conditioned media (Cho et al. 2009) suggesting a mechanism in which

Aβ can lead to mitochondrial fragmentation. This result, however, was challenged by

another group that showed that S-nitrosylation of DLP1 does not lead to an increase in its

GTPase activity as observed by the previous group, however they do show that DLP1

activity can be increased by NO through phosphorylation at a different site (Bossy et al.

2010). In this regard, it is of interest to note that Wang et al. demonstrated increased

phosphorylation of DLP1 at CDK1-dependent Ser585 sites (Wang et al., 2009). Also,

Yan et al. demonstrated that Aβ could lead to GSK3β activation which phosphorylates

DLP1 at Ser40/41 sites (Yan et al. 2015) and led to increased neuronal vulnerability.

Interestingly, Aβ mediated fragmentation could be caused by a Cdk5 regulated loss of

Mfn2 and Mfn1 (Park et al. 2015) setting up a complex stage of interactions between Aβ and mitochondrial dynamic proteins.

Emerging evidence also demonstrated mitochondrial dynamics abnormalities in

vivo in Drosophila and multiple mouse models of AD. Aβ-induced changes in mitochondrial dynamics and distribution are early events in vivo in Drosophila models.

Using primary neurons and brain tissue from several APP transgenic mice, Trushina’s

group found that mitochondrial impairments such as trafficking, morphology, and

respiratory activity, are observed preceding any plaque formation or memory deficits in

both Tg2576 and APP/PS1 mice (Trushina et al. 2012). The 3-D EM study revealed a

peculiar “beads-on-the-string” morphology which likely represented an excessive fission

process but stalled at the last step. Our group also demonstrated similar mitochondrial

fragmentation and ultrastructural damage as well as abnormal distribution in CRND8

mice preceding amyloid plaque deposition. Such mitochondrial fragmentation phenotype

was corroborated by an in vivo multiphoton imaging study, which depicted mitochondrial

fragmentation, loss of mitochondrial membrane potential, and loss of total number of

mitochondrial near surrounding of Aβ plaques in the brains of living mice (Xie et al.

2013).

Further efforts were taken to rescue AD-related deficits in animal models by

targeting mitochondrial fragmentation which adds more evidence of a critical role of

mitochondrial dynamic abnormalities in the course of AD. Our group demonstrated that

mdivi-1, an inhibitor of mitochondrial division, could rescue mitochondrial fragmentation

and axonal transport in primary neurons from CRND8 mice. More importantly, treatment

with mdivi-1 effectively rescued mitochondrial fragmentation in vivo in CRND8 mice.

By doing so, it also rescued mitochondrial dysfunction, oxidative stress and amyloid

pathology as well as cognitive deficits. Consistently, mdivi-1 could rescue the

mitochondrial dysfunction caused by Aβ oligomer treatment in hippocampal neurons

(Baek et al. 2017) and treatment of mdivi-1 in an APP/PS1 AD mouse model also

significantly improved the memory deficits.

Tau is necessary for Aβ-mediated neurodegeneration in AD, therefore it can be

assumed that tau also plays a role in mediating mitochondrial dysfunction in AD. An

abnormal interaction between hyperphosphorylated tau and DLP1 causes excessive

mitochondrial fission followed by degeneration in several AD mouse models as well as

AD brains (Manczak & Reddy 2012). Recently, they also showed that partial reduction in

DLP1 is protective against the hyperphosphorylated tau mediated mitochondrial

dysfunction (Kandimalla et al. 2016). The mechanism in which tau toxicity leads to mitochondrial dysfunction is unclear. Studies have shown that due to the interaction and stabilization of actin by tau, DLP1 does not properly associate with mitochondria resulting in elongation and neurotoxicity (DuBoff et al. 2012). Tau accumulation also

increases Mfn2/Mfn1 and OPA1 by decreasing ubiquitination in cell culture as well as

hippocampal neurons (Li et al. 2016) leading to mitochondrial elongation and subsequent

dysfunction. Interestingly these authors showed that mitochondrial elongation led to the

toxicity of tau, whereas above we see that Aβ-induced fragmentation is what leads to the

toxicity of Aβ. These contradicting observations shed light on the complex mechanisms

at play here. Perhaps there is a temporal relationship between tau- and Aβ-mediated

mitochondrial dysfunction that needs to be further investigated.

The transport of mitochondria plays an intricate role in mitochondrial distribution

and function since the ability to travel to areas of high energy demand is important, a

deficit in trafficking may lead to mitochondrial and neuronal dysfunction in AD (Wang et

al. 2009; Stokin et al. 2005). General signs of axonal defects such as axonal swellings and

spheroids precede the hallmark pathologies of Aβ or tau in both the AD postmortem brain

and AD mouse models (Stokin 2005). A pioneer study showed that brief exposure of hippocampal neurons to soluble Aβ halts mitochondrial motility and trafficking (Rui et

al. 2006). Later it was found that oligomeric Aβ reduces both retrograde and anterograde movement of mitochondria (Du et al. 2010; Calkins & Reddy 2011; Calkins et al. 2011).

Of note, it was found that reduced mitochondria in the axonal segments also led to a significant increase in the size of mitochondria suggesting that there is a relationship between mitochondrial transport and fission-fusion (Zhao et al. 2010).

Conclusion: Critical Role and Potential Mechanism of Mitochondrial Dynamics

Abnormalities in the Pathogenesis of AD

The central nervous system uses about 20% of the body’s daily energy intake while being only about 2% of the total body mass (Cunnane & Crawford 2014).

Mitochondria are almost exclusively the major source of energy production in the brain with glucose being their main energy substrate (Schönfeld & Reiser 2017). It stands to reason that any small irregularity in energy metabolism or production can lead to major adverse changes in such an important organ like the brain. Alzheimer’s Disease is a neurodegenerative disease affecting neurons of the cortical and medial temporal regions of the brain. It has been reported that one of the earliest pathological markers found in

AD is a reduced brain metabolic rate (Blass 2000). Along with these findings, others have reported damaged mitochondria output, increased ROS production, and increased mtDNA damage to be prevalent in AD (Gibson et al. 1998; Cottrell et al. 2001; Maurer

2000; Nagy et al. 1999; Hirai et al. 2001; Corral-Debrinski et al. 1994), leading us to the belief that mitochondrial health and neuron health go hand in hand and the inability of mitochondria to provide the necessary energy is what leads to neurodegeneration.

Mitochondria are very dynamic organelles as their morphology, distribution, and function are tightly regulated in order to maintain proper mitochondrial function. Mitochondrial fission and fusion play a key role in regulating not only morphology but also mitochondrial trafficking as well as function, and mutations in several of these fission/fusion regulators (i.e. Mfn2, OPA1, DLP1) leads to neurodegenerative diseases

(Kijima et al. 2005; Delettre et al. 2000; Waterham et al. 2007). Although there are no known mitochondrial mutations implicated in AD, mitochondrial dysfunction has been

shown to be a prominent and early feature in AD (Blass 2000; Swerdlow & Khan 2004;

Hirai et al. 2001). Our lab has previously shown mitochondria fission/fusion disruption

(fragmentation phenotype), altered expression of mitochondrial dynamic proteins (DLP1,

Mfn2, Mfn1, OPA1), and abnormal distribution (reduced neurite densities) in several

models of AD as well as AD patient brains (Wang et al. 2009; Wang et al. 2008). These

phenotypes are demonstrated in other APP- or Aβ-related in vitro models of AD as well as in vivo Drosophila and mouse models (Trushina et al. 2012; Xie et al. 2013; Wang et al. 2008; Wang et al. 2017). Collectively, these observations suggest that a disruption in mitochondrial dynamics plays a critical role in the pathogenesis of AD, yet whether or how mitochondrial fragmentation leads to neurodegeneration and what mechanisms lead to a reduction of fission/fusion proteins is not well understood which is the focus of my thesis. In this study, I used in vivo and in vitro models along with cellular and molecular

techniques to study how altered mitochondrial dynamics leads to neurodegeneration as

well as explore the mechanisms by which important mitochondrial dynamic proteins are lost. Given that mitochondrial dysfunction plays such a critical role in the pathogenesis of

AD, understanding the mechanisms in which mitochondrial function is altered may prove to be an important step in realizing a therapeutic treatment of AD.

Chapter 2. Mfn2 ablation causes an oxidative stress response and eventual neuronal death in the hippocampus and cortex

Abstract

Background: Mitochondria are the organelles responsible for energy metabolism and

have a direct impact on neuronal function and survival. Mitochondrial abnormalities have

been well characterized in Alzheimer Disease (AD). It is believed that mitochondrial fragmentation, due to impaired fission and fusion balance, likely causes mitochondrial dysfunction that underlies many aspects of neurodegenerative changes in AD.

Mitochondrial fission and fusion proteins play a major role in maintaining the health and function of these important organelles. Mitofusin 2 (Mfn2) is one such protein that regulates mitochondrial fusion in which mutations lead to the development of neurodegeneration.

Methods: To examine whether and how impaired mitochondrial fission/fusion balance

causes neurodegeneration in AD, we developed a transgenic mouse model using the

CAMKII promoter to knockout neuronal Mfn2 in the hippocampus and cortex, areas

significantly affected in AD.

Results: Electron micrographs of neurons from these mice show swollen mitochondria

with cristae damage and mitochondria membrane abnormalities. Over time the Mfn2

cKO model demonstrates a progression of neurodegeneration via mitochondrial

morphological changes, oxidative stress response, inflammatory changes, and loss of

MAP2 in dendrites, leading to severe and selective neuronal death. In this model,

hippocampal CA1 neurons were affected earlier and resulted in nearly total loss, while in

the cortex, progressive neuronal death was associated with decreased cortical size.

Conclusions: Overall, our findings indicate that impaired mitochondrial fission and

fusion balance can cause many of the neurodegenerative changes and eventual neuron

37 loss that characterize AD in the hippocampus and cortex which makes it a potential target for treatment strategies for AD.

Background

Alzheimer Disease (AD) is a multifactorial, age-related neurodegenerative disease of

the elderly that leads to progressive memory loss, impairments in behavior, language, visual-spatial skills and ultimately death. The disease is characterized by a progressive neuronal loss and accumulation of extracellular senile plaques composed of amyloid beta

(Aß) and intracellular neurofibrillary tangles (NFT) composed of hyperphosphorylated tau

in selective brain areas such as hippocampus and cortex(Mattson 2004). According to 2015

World Alzheimer’s Report, 47.5 million people had AD-related dementia worldwide,

including 5.4 million Americans, and projected the numbers to rise to 75.6 million by 2030

and to 131.5 million by 2050. Over 9.9 million new cases of AD-related dementia are

diagnosed every year worldwide(Salthouse 2004). Dementia has a huge economic impact

on our society and the estimated total healthcare cost of dementia worldwide in 2015 was

$818 billion. Currently, there are no drugs or agents available to treat or to prevent AD.

Several decades of intensive research have revealed that multiple cellular changes have

been implicated during the course of AD including widespread oxidative damage,

extensive neuroinflammation, and aberrant cytoskeletal alteration among which

mitochondrial dysfunction is an early prominent feature in susceptible neurons in the brain

from AD patients and models of AD, and likely plays a critical role in the pathogenesis of

AD(Mattson 2004),(Swerdlow 2016; Wang et al. 2014). Indeed, a reduced rate of brain metabolism preceding functional impairment is one of the best documented abnormalities in AD which is likely due to deficiency in several key enzymes of oxidative metabolism including cytochrome oxidase (COX) that have been consistently demonstrated in

AD(Swerdlow 2016; Wang et al. 2014). However, mechanisms underlying mitochondrial

dysfunction in AD remains elusive.

Mitochondria are dynamic organelles that continuously fuse with each other to form

larger tubular networks and divide into smaller structures, a process regulated by the balance of fission/fusion machinery mainly involving several large GTPases: mitochondrial fission is regulated by cytosolic protein DLP1 which translocates to mitochondria during fission while mitochondrial fusion is regulated by mitofusin 1 and 2

on the outer mitochondrial membrane and OPA1 on the inner mitochondrial

membrane(Mishra & Chan 2014). The delicate balance between mitochondrial fission and

fusion is crucial in the maintenance of healthy population of mitochondria and proper

mitochondrial distribution, morphology and function, disruption of which causes human

diseases, especially neurological diseases(Mishra & Chan 2014). Increasing evidence

suggested that an abnormal mitochondrial dynamics is likely involved in the mitochondrial

structural damage and dysfunction in AD: several groups demonstrated that overexpression

of familial AD-causing amyloid precursor protein (APP) mutants or exposure to soluble

oligomeric Aβ caused changes in the expression of mitochondrial fission and fusion

proteins and profound mitochondrial fragmentation in neuronal cells which led to

ultrastructural damage to mitochondria and mitochondrial dysfunction along with neuronal

deficits such as synaptic abnormalities in vitro(Wang et al. 2008; Manczak et al. 2010; Du

et al. 2010; Wang et al. 2009; Calkins & Reddy 2011; Wang et al. 2010). Aβ-induced changes in mitochondrial dynamics and distribution are early events in vivo in Drosophila models(Zhao et al. 2010; Iijima-Ando et al. 2009). Abnormal mitochondrial distribution and round, swollen, and damaged mitochondria consistent with enhanced fragmentation

are also documented in AD mouse models(Trushina et al. 2012; Baek et al. 2017; Wang et

al. 2017). Fibroblasts from AD patients and AD cybrid cell models also demonstrated

abnormal mitochondrial dynamics and dysfunction(Wang et al. 2008; Gan et al. 2014).

Indeed, ultrastructural deficits and abnormal distribution of mitochondria were evident in

pyramidal neurons in AD brain(Wang et al. 2008). Studies from several groups consistently demonstrated a significantly reduced expression of large GTPases involved in fusion (i.e.,

Mfn1, Mfn2 and OPA1) in the brain of AD patients, implicating that mitochondrial fragmentation, largely due to disrupted mitochondrial fusion, likely occurs in the susceptible neurons in AD brain(Wang et al. 2009; Calkins et al. 2011; Manczak et al.

2011). Mitofusin 2 is expressed in many types of cells and tissues. Thus far, genetic mutations were found and reported to be associated with Charcot-Marie-Tooth (CMT) disease, the most common inherited neurological disorders, accounting for up to 20-30% of all axonal CMT type 2 cases, to a lesser degree, also be associated with optic atrophy, clinical signs of first motor neuron involvement, and early onset stroke. Although Mfn2 expression levels were found to be reduced in AD patient brains and AD mice, there are no mutations yet reported to be associated with AD or PD patients. However, whether disruption of mitochondrial fusion causes mitochondrial deficits and neurodegeneration in

AD-afflicted brain areas has not been determined. To address the causal role of disrupted mitochondrial fusion in neurodegeneration and other AD-related deficits in AD-afflicted brain areas, we disrupted mitochondrial fusion by knocking out Mfn2 in the hippocampus and cortex and found that ablation of Mfn2 caused neuronal degeneration in vivo in a temporal order starting with significant mitochondrial fragmentation and dysfunction and increased oxidative stress, followed by cytoskeletal alterations and inflammatory response

41 that culminates in eventual neuronal death, all features characterizing AD, thus establishing the critical role of mitochondrial fragmentation in mitochondrial dysfunction and neuronal degeneration and the pathogenesis of AD.

Methods

Transgenic mice

Experimental mice were generated by crossing B6.Cg-Tg(Camk2a-cre)T29-1Stl/J mice (Jackson Laboratory) with B6.129(Cg)-Mfn2tm3Dcc/J mice (Mfn2loxP). PCR of genomic DNA was used for routine genotyping of offspring to detect the presence of the

CAMKCre gene and the Mfn2loxP gene. We chose these mice since the Cre-recombinase was shown to be expressed in the forebrain and hippocampus, highest in the CA1 layer.

Mice were housed under standard conditions and all animal studies were performed following an approved protocol through the Case Western Reserve University IACUC board. At various ages, mice were perfused with saline and brain tissue collected and bisected with one half frozen (with brain areas cortex, hippocampus, cerebellum and brainstem area separated) and the other half fixed in 10% buffered formalin. For this constitutive expression study, Mfn2KO mice from ages 4, 8, 12, 18, 28, and 78 weeks (at least n=3 per age group) and age-matched control mice (no Cre+, n=3) were collected. In addition, hemizygous mfn2 KO were examined at 8 and 12 weeks, and 18 months.

PCR

To confirm recombination of the floxed gene, DNA was extracted from frozen samples of cortex, hippocampus, cerebellum and brainstem using the Wizard genomic

DNA kit. PCR was performed using primers designed to amplify a 240bp band only after

Cre recombination has occurred.

Immunoblotting

For Western blot analysis frozen dissected hippocampal and cortical tissues were individually homogenized in 10X volume Cell lysis buffer (Cell signaling) with added

43 protease and phosphatase inhibitors (Roche) and centrifuged at 14,000 g for 20 minutes in a refrigerated microcentrifuge. Protein concentration of the supernatant was determined using BCA protein assay. Proteins, 10 μg per lane (specified in figure legends), were resolved with SDS-PAGE and transferred to Immobilon (Millipore). Blots were blocked in 10% nonfat milk for 1 hour and probed with primary antibodies overnight. After washing 5X in TBS-tween, HRP-conjugated secondary antibodies (Cell

Signaling) were applied for 1 hour, rinsed again and bands detected using ECL (Santa

Cruz or Millipore). Loading control was anti-actin (Millipore). Bands were quantified using QuantityOne (Biorad) or ImageJ.

Immunohistochemistry

Formalin fixed brain samples were embedded in paraffin and 6 μm sections were cut using a microtome and placed on coated slides. Immunohistochemistry was performed using the peroxidase anti peroxidase method as previously described(Wang et al. 2016).

Briefly, sections were deparaffinized in 2 changes of xylene, dehydrated through descending series of ethanol, and finally into Tris buffered saline (TBS: 50 mM Tris, 150 mM NaCl, pH=7.6). Endogenous peroxidase was removed with a 30-min incubation in

3% H2O2. Antigen retrieval using citrate buffer and pressure cooking (Biocare) was performed for most IHC experiments. After blocking for 30 min in 10% normal goat serum (NGS), primary antibodies were applied and incubated overnight at 4°C. Species specific secondary antibodies and PAP complexes were applied and after 3 changes of tris buffer, the sections were developed with DAB (Dako) and slides were rinsed in dH2O, dehydrated and coverslipped with Permount. Antibodies used were directed against Cre-recombinase (Millipore), GFAP (Invitrogen), MAP2 (Millipore), NeuN

(Millipore), COXI (Molecular Probes), HO-1 (Enzo), iba1 (Wako), AT8 (Thermo

Fisher), and mitochondria complex cocktail (Abcam).

To detect protein carbonyls, the dinitrophenol binding -assay was also performed.

After tissue sections were rehydrated, 50 ul of DNP solution was applied (20mM DNP,

0.5% TFA, 92.5% DMSO) and incubated for 15 min at 37° C. The sections were rinsed with changes of acetic acid until no yellow color could be further removed. After rinsing thoroughly with water and then with TBS, sections were blocked and DNP detected with rabbit antibody to DNP (1/10,000). For western blot analysis, samples were first denatured and then treated with DNP solution following manufacturer’s instructions

(Millipore), sample buffer directly added and the entire sample run on the gel.

TUNEL method was used to label cells undergoing apoptosis following manufacturer’s instructions (Roche). After TUNEL method, the sections were stained with DapI and mounted with Fluoromount (SouthernBiotech) and FITC-labelled apoptotic cells were imaged on a Zeiss Axiophot.

Image Quantification

Images were acquired on a Axiophot with a axiocam (Zeiss). Quantification was performed on these images using Axiovision software and either the number of cells stained per area, or density of stained structures was determined using Axiocam image analysis software. To measure hippocampal size, H&E stained sections from brain areas collected from Bregma 1.08-1.8, representing the complete hippocampus and similar ventricle presentation were imaged and measured in a blinded fashion. To measure total cortical size, sections from similar levels were imaged at the same magnification, the cortical area cropped using Photoshop and the area was quantified.

Electron Microscopy

For electron microscopic analysis, brain samples from control and KO mice were collected and fixed as previously described following brain dissection(Wang et al. 2017).

Brain slices of about 1 mm thick were made and small areas of the CA1 region of the hippocampus and cortex were sampled and embedded in Epon. Semithin sections were prepared and stained with toluidine blue to clearly note the CA1 regions. Images of pyramidal neurons from the CA1 region or cortex with initial segment of axon visible were obtained. Mouse genotype was blinded to the electron microscopist. Mitochondria parameters were quantified using Image J and included aspect ratio (length/width) and size (area in μm2).

Mitochondrial oxygen consumption measurement

The real-time measurement of oxygen consumption rate (OCR) in synaptic mitochondria in synaptosomes was performed using the Seahorse XF24 Analyzer

(Seahorse Bioscience, North Billerica, MA), according to the manufacturer's instructions.

If needed, ATP synthase inhibitor oligomycin (1 μM), uncoupler FCCP (4 μM) and complex I inhibitors antimycin A (1 μM) and rotenone (1 μM) were injected sequentially.

After measurement, cells and synaptosomes were lysed and OCR data was normalized by total protein as previously described(Wang et al. 2016).

Results

Conditional knockout of Mfn2 in the forebrain

Mfn2 levels are significantly reduced in AD and likely contribute to increased

mitochondrial fragmentation(Wang et al. 2009; Manczak et al. 2011). To assess the

causal relationship between disrupted mitochondrial fusion and AD-related deficits in the

brain areas affected by AD, we utilized a genetic approach by crossing the Mfn2

conditional knockout mice (Mfn2loxP/loxP) with CaMKII-Cre mice, in which the Cre recombinase is expressed in the forebrain and result in specific ablation of Mfn2 in selective neurons in the hippocampus and cortex, brain areas that are heavily afflicted in

AD. Genomic DNA PCR analysis confirmed Cre-mediated recombination and excision of floxed Mfn2 (Fig. 2.1A) in the hippocampus and cortex but not in the cerebellum in the homozygous knockout mice (CaMKII-Cre+/-/Mfn2loxP/loxP, hereafter referred to as

Mfn2 cKO). Western blot revealed that Mfn2 protein levels were significantly reduced in

the hippocampus of the homozygous Mfn2 cKO mice, compared to control mice (Fig.

2.1B), since the age of 8 weeks (Fig. 2.1C). There was a trend towards decreased Mfn2 in

the cortex at ages of 8 weeks which became significant from 12 weeks of age and

thereafter. (Fig. 2.1C). To confirm the specific expression of Cre and the loss of Mfn2

protein in the selective neurons of the hippocampus and cortex, we performed

immunocytochemical study in the 8 week old mice brains (Fig. 2.1D). Consistent with

previous studies using CaMKII-Cre mice to ablate the floxed target gene(Shields et al.

2015), there was a selective expression of Cre and loss of the Mfn2 protein (Fig. 2.1D) in

the cortical neurons and in pyramidal neurons in the CA1 region but not CA2 region of

the hippocampus. We also determined the expression of other fission and fusion proteins

in the Mfn2 cKO mice by western blot and found no significant changes in the expression

of DLP1 or OPA1 in the Mfn2 cKO mice at 8 weeks of age (Fig. 2.2). The Mfn2 cKO

mice develop and grow normally and display normal fertility with no gender differences.

Mfn2 knockout caused mitochondrial ultrastructural deficits and abnormal distribution

To visualize the impact of disruption of mitochondrial fusion by Mfn2 ablation on

mitochondrial morphology and distribution in the hippocampus, electron microscopy was

performed to examine mitochondria in the CA1 pyramidal neurons. Tubular

mitochondria with the regular, accordion-like folds of cristae without any notable abnormalities were found in the CA1 pyramidal neurons in both the youngest (i.e., 4 weeks) (not shown) and the oldest littermate control mice examined (i.e., 28 weeks) (Fig.

2A). Similarly, at the age of 4 weeks, no obvious abnormalities in the appearance of

mitochondria were noted in the Mfn2 cKO mice (Fig. 2.3A). However, dramatic changes in mitochondrial morphology and cristae organization were noted in the CA1 pyramidal neurons of Mfn2 cKO mice by the age of 8 weeks (Fig. 2.3A): mitochondria appeared

rounder and swollen with broken cristae, many of them also exhibiting multilamellar

appearance and vacuolation. It became more severe in the neurons of 28 week old Mfn2

cKO mice where extreme loss of internal cristae structure was frequently seen (Fig.

2.3A). Quantification of all mitochondria in 4-5 neurons per mouse find that the average mitochondrial aspect ratio (length/width of each mitochondria) was unchanged in 4 week

Mfn2 cKO mice but became significantly decreased in 8 week and 28 week Mfn2 cKO

mouse, compared to control mice (Fig. 2.3B). Interestingly, in 28 week Mfn2 cKO mice,

mitochondrial size was significantly increased and the mean size almost doubled

compared to the control mice (Fig. 2.3C).

Additionally, mitochondria in neurons from control mice are generally distributed evenly along the entire neuronal processes, however, mitochondria in the processes of 8 week Mfn2 cKO mice were far less numerous, demonstrating abnormal mitochondrial distribution. This abnormal distribution of mitochondria is further exacerbated in the 28 week Mfn2 cKO mice (Fig. 2.4A). Consistently, immunohistochemistry using antibodies against mitochondria complex proteins demonstrate specific changes in the staining pattern that also confirmed abnormal mitochondrial distribution in Mfn2 cKO mice: mitochondria are distributed throughout the neuronal cell body and along the processes in control mice at all ages examined, but instead accumulate mainly in cell body leaving processes largely devoid of mitochondria staining in both the hippocampus CA1 neurons and cortical neurons in as early as the 8 week old Mfn2 cKO mice (Fig. 2.4B,C).

Mfn2 knockout caused mitochondrial dysfunction

We then characterized whether and how mitochondrial function is affected in Mfn2 cKO mice. To examine changes in the expression of proteins involved in electron transport chains, we performed western blot analysis of the hippocampal tissues from the control and Mfn2 cKO mice. Starting at 8 weeks old and continuing with age, a significant loss of complex I is seen in the hippocampus of the Mfn2 cKO mice (Fig.

2.5A,B). There is a decrease trend in Complex II which becomes significant at 12-18 weeks of age. No changes were found in either Complex III or V even by 28 weeks (data not shown). To determine the impact of Mfn2 cKO on mitochondrial function, mitochondrial respiration was determined in freshly isolated synaptic mitochondria from hippocampus of 8 weeks old Mfn2 cKO mice and their littermate controls using a

Seahorse XF24 extracellular flux analyzer. Both basal and maximal oxygen consumption

rate (OCR) along with spare respiratory capacity (difference between basal and maximal

respiratory capacity, critical for neuronal bioenergetics under stress) were significantly reduced in the synaptic mitochondrial from Mfn2 cKO mice as compared with littermate control mice (Fig. 2.5C-E,G). Respiration control ratio calculated from OCR measurement after FCCP and oligomycin treatments was also significantly decreased in

Mfn2 cKO mitochondria but mitochondrial coupling efficiency remains unchanged (Fig.

2.5F,H).

Mfn2 cKO caused severe neurodegeneration in the hippocampus and cortex

H&E stains revealed apparent hippocampal neuronal degeneration in the CA1 area with aging in Mfn2 cKO mice compared with age-matched littermate control mice (Fig.

2.6A,B): Neurons in the hippocampus in Mfn2 cKO mice up to 8 weeks appeared normal; starting at 12 weeks of age, many of the CA1 neuronal nuclei exhibited a shrunken appearance although no significant changes in the number of neurons were noted in Mfn2 cKO mice. Mfn2 cKO mice at ages 18 weeks and 28 weeks demonstrated continued loss of hippocampal neurons, yet the dentate gyrus and CA2 neurons remained intact across all ages. The specific degeneration of neurons in the Mfn2 cKO mice was demonstrated by NeuN staining (Fig. 2.6C). Neuronal number in the hippocampus CA1 and CA2 regions quantified using the sections stained for NeuN revealed that the number of pyramidal neurons in the CA2 region remained constant with age (not shown), yet the

number of pyramidal neurons in the CA1 region showed a trend towards decrease at 12

weeks of age which only became significant starting from age 18 weeks (Fig. 2.6D).

Indeed, the hippocampus size was also reduced significantly with age in the Mfn2 cKO mice but not in littermate control mice (data not shown).

Examination of the entire cortical region of sections stained with NeuN reveals severe

shrinkage of the cortex in the Mfn2 cKO mice (Fig. 2.6C). In fact, quantification of the

entire cortical area shows a slight but significant decrease in the cortical size in control

mice with aging which became much more severe and significant in the Mfn2 cKO mice

(Fig. 2.6E). Higher magnification images of the cortical region proximal to the

hippocampus using Nissl staining showed the neuronal layers became disorganized with

many nuclei disfigured and less uniform at 28 weeks of age (Fig. 2.7). The total number

of NeuN-positive neurons in the cortex decreased significantly since 18 weeks of age and further decreased at 28 weeks of age (Fig. 2.6D).

TUNEL staining revealed apoptosis only at 18 weeks of age (Fig. 2.6G) in both the

CA1 region and in the cortex of the Mfn2 cKO mice. No apoptotic cells were evident in

any of the 12 week old Mfn2 cKO mice. DapI staining reveals some neuronal loss in the

CA1 in the 12 week mice compared to the control mice (not shown) or 4 week old Mfn2

cKO mice (Fig. 2.6G), which correlates with the appearance of shrunken nuclei seen with

H&E staining shown in Fig. 2.6B at this age.

Mfn2 knockout caused increased oxidative stress

Widespread oxidative stress is an early and prominent feature of AD which is

believed plays a role during neurodegeneration(Nunomura et al. 2012).

Immunocytochemical studies revealed increased protein oxidation as demonstrated by

enhanced DNP immunostaining suggestive of increased protein carbonyls as early as 8

weeks in the Mfn2 cKO mice which became more prominent with age (Fig. 2.8A).

Indeed, by a DNP Oxyblot, protein carbonyls were significantly increased at 8 weeks in

the hippocampus (Fig. 2.8B,C). These findings are corroborated through immunostaining

of brain tissue for another oxidative marker, the antioxidant enzyme hemoxygenase-1

(HO-1): The pyramidal neurons demonstrate progressively increased levels of HO-1 in

the Mfn2 cKO hippocampus and cortex since the age of 8 weeks (Fig. 2.8D,E).

Mfn2 knockout caused increased neuroinflammation

Increased inflammation is another characteristic of AD(Yin et al. 2018). Increased

gliosis was found in the specific brain regions that displayed the neuronal degeneration

and loss in older Mfn2 cKO mice (Fig. 2.9). Some increased GFAP-immunostaining

began to appear in the CA1 region of hippocampus of Mfn2 cKO mice starting at 8

weeks of age which became prominent and significantly increased since 12 week of age

(Fig. 2.9B and Fig. 2.10). The CA2 region remained spared from astrocyte accumulation in the Mfn2 cKO mice even at age 28 weeks (Fig. 2.9B). In the cortex, GFAP-positive astrocytes accumulate progressively with age appearing first in neuronal layer V/VI at 12 weeks and then encompassing all cortical neuronal layers by 18 weeks of age (Fig. 2.9D).

Microglia, stained using antibody against iba1, were also activated in the Mfn2 cKO mice at 12 weeks of age, with Type IV amoeboid microglia appearing in the peripheral

CA1 areas just adjacent to the subiculum, then encompassing the remaining CA1 by 18 weeks of age, (Fig. 2.9C) and in the cortical areas (Fig. 2.9E), which correlates with loss of neurons detected with antibody NeuN (Fig. 2.6A). By 28 weeks of age, activated microglia are absent in the CA1 (Fig. 2.9C,E), likely since most neurons are gone, yet

GFAP-positive astrocytes continue to accumulate (Fig. 2.9B,D). No apparent microglia activation is observed in the CA2 regions at all ages examined.

Mfn2 knockout caused cytoskeletal alterations

Cytoskeletal changes were also apparent in the neurons where Mfn2 was ablated with

age (Fig. 11). In the hippocampus, MAP2 is present in CA1 neuronal dendrites, with a

diffuse stain dispersed evenly in the cell bodies in control mice through all ages

examined. Similar staining pattern of MAP2 was noted in Mfn2 cKO mice only up to the

age of 8 weeks (Fig. 2.11A). By 12 weeks, however, the cellular distribution of MAP2 is

changed dramatically in the CA1 region, such that there is a significant loss of dendritic

staining, which correlates with increased localization of MAP2 in the cell body in Mfn2

cKO mice. This pattern remains until age 18 weeks, when very little MAP2

immunoreactivity remains in either the process or cell bodies (Fig. 2.11A). Throughout the cortex, with increasing age, neurons in the Mfn2 cKO animals showed increasingly stronger cell body staining, with concomitant thickening of the processes (Fig. 2.11B).

Higher levels of hyperphosphorylated tau, stained using the AT8 antibody, are found in the CA1 neurons in 8 week Mfn2 cKO mice compared to the control mice. The AT8 immunostaining becomes more prominent in 12 week cKO mice with increased staining in the neuronal processes (Fig. 2.11C).

Discussion

Studies from multiple groups suggest that mitochondrial fragmentation largely due to

disrupted mitochondrial fusion contributes to mitochondrial dysfunction and neuronal deficits in AD. To gain a better understanding of whether and how mitochondrial

fragmentation causes AD-related deficits in the hippocampal and cortical neurons, in this

study, we developed a transgenic mouse model of disturbed mitochondrial fusion in the pyramidal neurons of hippocampus and cortex by tissue-specific knockout of Mfn2.

Interestingly, we found that ablation of Mfn2 caused a series of pathological events including mitochondrial damage and dysfunction, oxidative stress, inflammatory response, and cytoskeletal changes that eventually led to significant neurodegeneration in the hippocampus and cortex in vivo.

In our study, mitochondrial fragmentation became apparent in the CA1 hippocampal neurons since 8 weeks of age, which was accompanied by increased ultrastructural damage to mitochondria as reflected by the loss of integrity of the internal structures and the apparent swollen appearance of mitochondria. Interestingly, increased oxidative stress and impaired mitochondrial function such as reduced expression of OXPHO proteins and decreased mitochondrial respiratory parameters were also observed in the Mfn2 cKO mice at this same age. While the causal relationship between mitochondrial dysfunction and damage, and oxidative stress remains to be teased out, it was suggested that excessive mitochondrial fission leads to reduced mtDNA copy number(Chen et al. 2010) which could lead to reduced expression of essential OXPHOS proteins as we confirmed by the western blot analysis and cause mitochondrial dysfunction. Prior studies from multiple groups demonstrated that excessive fission or fusion caused deficits in complex assembly(Liu et

54 al. 2011; Zhou et al. 2017; Cogliati et al. 2013) likely through changes in the cristae shape and curvature or other uncharacterized mechanism that also negatively impacted mitochondrial function. Indeed, dramatic changes in the cristae organization was noted in the 8 weeks old animals. Mitochondrial dysfunction unavoidably causes increased oxidative stress which could damage mitochondrial ultrastructure and further exacerbate mitochondrial dysfunction. The progression of mitochondrial ultrastructural damage from relatively mild changes such as multilamellar appearance and vacuolation with partially broken cristae in Mfn2 cKO at 8 weeks of age to more severe appearance such as the extreme loss of internal cristae structure and significantly swollen mitochondria likely reflected the accumulation of oxidatively damaged mitochondria along the course. It is of importance to note that no mitochondrial defects were noted in the brain of the Mfn2 cKO mice at 4 weeks of age which excludes the potential complication due to developmental abnormalities. These results clearly demonstrated that Mfn2 ablation-induced mitochondrial fragmentation caused mitochondrial structural damage and dysfunction in the hippocampus and cortex in vivo.

The most striking change in the Mfn2 cKO mice is the apparent neurodegeneration in the hippocampal area at 18 weeks of age where near 90% of the CA1 hippocampal neurons were wiped out. While this extreme loss of neurons was not apparent in individual fields of the cortical area, taken together with the progressive shrinkage of the entire cortex with age, significant cortical neuron loss was apparent at 18 weeks of age when around 25% neurons were lost. Indeed, TUNEL assay did reveal that these neurons die by apoptosis in both the hippocampus and cortex at 18 weeks of age. Significant cortical neuronal loss continues to progress and reached around 50% at 28 weeks of age. It was noted that

disabling mitochondrial function could produce the same pathological changes. For

example, malonate treatment, which inhibits succinic dehydrogenase, resulted in strikingly

similar hippocampal neurodegeneration(Davolio & Greenamyre 1995). Given that mitochondrial dysfunction far precedes neurodegeneration in this Mfn2 cKO model, our results clearly demonstrated that disrupted mitochondrial fusion could lead to neurodegeneration in AD-affected brain areas through mitochondrial dysfunction.

Interestingly, Mfn2 ablation induced neurodegeneration in the hippocampus and cortex

is preceded by a series of pathological events in temporal order that may suggest a causal

relationship between these events. Increased oxidative stress was observed in the Mfn2

cKO mice at the age of 8 weeks. It must be noted that oxidative stress is also a prominent

and early feature of AD(Zhu et al. 2005). For example, advanced glycation end-products,

lipid peroxidation, protein nitration and carbonyl formation have all been found to be

increased in the brain of AD(Zhu et al. 2005). Further induction of HO-1 was determined

to be a relatively early neuronal response as its appearance co-localized with the Alz50 tau

epitope in degenerating neurons in AD(Takeda et al. 2000). At 12 weeks of age,

cytoskeletal and neurofibrillary changes became apparent in the Mfn2 cKO mice. As

temporally later events, cytoskeletal and neurofibrillary changes likely lie downstream of

mitochondrial dysfunction and/or increased oxidative stress. In this regard, ample evidence

demonstrated that oxidative stress could impact the posttranslational modification and lead

to increased proteolysis of microtubule associated proteins, reduce the ability of

microtubule to polymerize and cause severing of actin microfilaments and thus impair

cytoskeletal structure in both neuronal and non-neuronal cells(Kwei et al. 1993; Akulinin

& Dahlstrom 2003; Pang et al. 1996). Soluble Aβ oligomers has been shown to cause

proteolysis of microtubule associated proteins including (MAP2) during apoptosis(Fifre et

al. 2006). Similarly, oxidative stress causes increased tau phosphorylation and promotes

the conformational changes and fibrillation/aggregation of tau protein which leads to AD- related neurofibrillary changes(Su et al. 2010; Takeda et al. 2000; Q. Liu et al. 2005; Pérez et al. 2000). An inflammatory response was also observed in the Mfn2 cKO mice at 12 weeks of age which likely is also caused by mitochondrial dysfunction and/or oxidative stress. The inflammatory response, as shown by increased activation of microglia and astrocytes, precedes the observed neurodegeneration at 18 weeks of age. There are many studies that suggest inflammation as an early sign of AD disease progression. An inappropriate immune response may promote AD by increasing the production of Aβ and reducing removal of amyloid plaques by microglia (Paolicelli et al. 2017). Overall, our study demonstrated that mitochondrial fragmentation caused by disruption in mitochondrial fusion could initiate a cascade of abnormal changes that are relevant to the important pathological changes during the course of AD and lead to neurodegeneration in the hippocampus and cortex in vivo.

In addition to mitochondrial fragmentation and ultrastructural damage, we found an abnormal distribution of mitochondria, especially in the neuronal process, in the pyramidal neurons of Mfn2 cKO mice months before neurodegeneration. Loss of Mfn2 leads to decreased number of mitochondria in the axon and dendritic processes. In this regard, it is important to note that such an abnormal distribution of mitochondria was also noted in the brain of AD patients(Wang et al. 2008) and AD animal models(Wang et al. 2017). This

observation replicated prior findings where Mfn2 deletion also caused a prominent defect

in mitochondrial content and transport in the processes of dopaminergic neurons which set

DA neurons to degenerate 1-2 months later in retrograde manner(Lee et al. 2012; Pham et

al. 2012). While it is not clear whether such abnormal mitochondrial distribution contribute

to neurodegeneration in the hippocampal and cortical neurons in Mfn2 cKO mice since

some neuron population could survive almost complete loss of mitochondria from its

processes(Berthet et al. 2014), it could contribute to neuronal dysfunction such as synaptic

deficits of importance to AD as we demonstrated before(Wang et al. 2008). Although

altered expression of various mitochondrial fission/fusion proteins could differentially

impact mitochondrial transport(Wang et al. 2009), Mfn2 may impact mitochondrial transport through its direct interaction with Miro and Milton(Misko et al. 2010), adaptor proteins crucial for kinesin-mediated transport of mitochondria.

As we demonstrated in this study, CA1 neurons are susceptible to Mfn2 loss. However, these neurons are resistant to DLP1 loss(Shields et al. 2015). It is of interest to note that, in the contrary, dopaminergic neurons and Purkinje cells are susceptible to both the loss of

Mfn2 and DLP1(Pham et al. 2012; Berthet et al. 2014; Chen et al. 2007; Wakabayashi et al. 2009; Lee et al. 2012), respectively. It is believed that different cells, including various neuron populations, have very different mitochondrial morphology according to their specific metabolic needs, and hence possess unique balance on the regulation of mitochondrial dynamics. Therefore, such a difference in the tolerance of loss of Mfn2 and

DLP1 suggests that, comparing to other neuron populations, CA1 neurons more critically depend on mitochondrial fusion than fission for their function and survival which makes mitochondrial fusion a better target for restoring mitochondrial dynamic balance in AD.

Alternatively, Mfn2 has been implicated in the regulation of mitochondrial properties besides fusion such as mitochondrial transport(Misko et al. 2010). This could point to a

58 difference in mitochondrial distribution to explain the preferential requirement of Mfn2 in the hippocampus for neuronal survival. However, this appears unlikely since studies of

DLP1 loss in dopaminergic neurons show a similar abnormality in mitochondrial movement suggesting that DLP1 may also play a role in mitochondrial transport(Berthet et al. 2014). In this regard, the loss of DLP1 in the hippocampus increased the distance between mitochondria in the dendritic processes which was well tolerated, although overall content and number were unchanged(Shields et al. 2015).

Overall, in this study we demonstrated that Mfn2-ablation induced mitochondrial fragmentation led to neurodegeneration through mitochondrial dysfunction and increased oxidative stress and a series of event in a strict temporal order in mice and all these pathological changes are also characteristics seen in AD during the course of disease. These results do not necessarily suggest that the pathological events occur at a similar temporal order in the AD brain, however, they suggest that disrupted mitochondrial dynamics and mitochondrial dysfunction could contribute to these pathological events in AD.

Figure 2.1 Cre-mediated ablation of Mfn2 expression in the hippocampus and cortex of Mfn2 cKO mice. a Cre-mediated recombination and excision of floxed Mfn2was analyzed by genomic DNA PCR. Cre- mediated ablation of Mfn2 was found in the hippocampus (Hip) and cerebral cortex (Cx) but not in cerebellum (Cb). Western blot (b, representative from animals of 8–12 weeks of age) and quantification analysis (c) of brain homogenates found the Mfn2 protein levels were reduced in Mfn2 cKO mice (N = 7) compared to control mice (N = 6) in both the hippocampus and the cortex. Actin was used as an internal loading control. Data are means ± SEM, student’s t-test, *P < 0.05, ***P < 0.001. d Mfn2 immunostaining of brain sections from 8-week-old control and Mfn2 cKO mice showed a loss of Mfn2 protein in both cortical (boxed area was enlarged in the upper-right corner) and CA1 hippocampal neurons but not CA2 hippocampal neurons (boxed areas were enlarged below)

Figure 2.2 Representative western blot and quantification analysis of brain homogenates found the DLP1 and OPA1 protein levels were unchanged in Mfn2 cKO mice (8 weeks of age) compared to control mice in both the hippocampus and the cortex. GAPDH was used as an internal loading control (N=3/group, data represent mean ±SEM, Student’s t-test, *P < 0.05).

Figure 2.3 Mfn2 ablation caused mitochondrial fragmentation and ultrastructural damage in the hippocampus in vivo as evidenced by electron microscopic analysis. a Representative electron micrographs of CA1 neuron from hippocampus of control (28- week-old) (upper left) and Mfn2 cKO at different ages as indicated. b Quantification of aspect ratio demonstrated significant mitochondrial fragmentation in 8 and 28 week old Mfn2 CKO mice. c Quantification of mean area of mitochondria demonstrated significantly enlarged mitochondrial size in 28 week old Mfn2 cKO mice compared to control mice. Data are means ± SEM Student’s t-test, *P < 0.05

Figure 2.4 Mfn2 ablation caused abnormal mitochondrial distribution in vivo. a Representative electron micrographs of a segment of axon of CA1 neuron from hippocampus of control and Mfn2 cKO mice. Immunohistochemistry using OXPHOS cocktail antibody to stain mitochondria finds many neurons in the hippocampus (b) and frontal cortex (c) exhibiting staining throughout the cell body and processes in the control mice. However, at as early as 8 weeks, a striking loss of neuronal process immunolocalization in the CA1 and cortical neurons occurs and becomes more apparent with aging

Figure 2.5 Mfn2 ablation caused mitochondrial dysfunction in the brain of Mfn2 cKO mice. Representative western blots (a) and quantification analysis (b) showed changes in the protein levels of key subunits of respiratory complex I (anti-NDUFB8), and II (anti- SDHB) in the hippocampal and cortical tissues of Mfn2 cKO mice since 8 week of age. (c) Respiratory activity of synaptic mitochondria freshly isolated from hippocampus of 8- week-old Mfn2 cKO mice and control mice was analyzed by Seahorse XF Assay. Oxygen consumption (OCR) rate was measured before and after sequential exposure to oligomycin (inhibits ATP synthase, blocks oxygen consumption related to ATP synthesis), FCCP (uncoupler to assess maximal OCR), and antimycin A/rotenone (blocks electron flux through both complex I and II). Quantification of basal (d) and maximal (e) OCR of control and Mfn2 cKO mice. Respiration control ratio (F) (OCRFCCP/OCROligomycin), Spare respiratory capacity (g) (OCRFCCP-OCRBasal) and Coupling efficiency (H) ([OCRBasal -OCROligomycin]/ OCRBasal) were calculated after subtracting the non-mitochondrial respiration (OCRAntimycin A/rotenone). Data are means ± SD of 3 mice. Statistics: Student’s t test. *p < 0.05 and **p < 0.01 compared with control

Figure 2.6 Mfn2 ablation caused neurodegeneration in the hippocampus and cortex in vivo. a Representative H&E stains of the hippocampus revealed an obvious pattern of hippocampus neuronal degeneration in the CA1 area of Mfn2 cKO mice with aging, yet the dentate gyrus and CA2 remained intact across all ages. b Enlarged picture of H&E staining of the CA1 neurons revealed a shrunken appearance of nuclei of many CA1 neurons in the Mfn2 cKO mice starting at 12 weeks of age. c Using low magnification images immunostained using NeuN, the cortical area also becomes shrunken with age (c) and quantification reveals this is a significant correlation (F, p < 0.001). d Enlarged picture of the boxed areas in (c) of the NeuN staining of the CA1 neurons. Quantification revealed a significant loss of neurons in both the cortex and CA1 of the hippocampus at 18 weeks (e). Neuronal apoptosis detected by TUNEL assay was only seen at 18 weeks in both the hippocampus and cortex (g). No apoptotic cells were present in any of the 4 or 12 week old cKO mice, but the number of CA1 neurons is reduced in the 12 week old mice compared to the 4 week old mice seen in the DapI images, reflecting the cellular changes noted by H&E staining at 12 weeks of age

Figure 2.7 Nissl staining of the entire cortex showing neuronal layers I-VI (upper panels) and higher magnification of area encompassing layers III/IV (lower panels). Dashed line represents lower boundary of cortical neurons. Cortical shrinkage is apparent, neuronal layers become less distinct, and nuclei become disfigured and less uniform with age in the cKO mice.

Figure 2.8 Mfn2 ablation caused increased oxidative stress in the hippocampus and cortex in vivo. A Representative of DNP adduction immunochemical assay to detect protein oxidation in the hippocampus of Mfn2 cKO mice and control mice. Representative western blots (b) and quantification analysis (c) showed significant increase in protein oxidation in 8-week-old Mfn2 cKO mice compared to age-matched control mice using the DNP assay. Actin was used as an internal loading control. (Data represent mean ± SEM, Student’s t-test, *P < 0.05). d-e Representative immunohistochemistry of anti-HO-1, an inducible antioxidant enzyme, in the hippocampus (d) and cortex (e) of Mfn2 cKO mice and control mice

Figure 2.9 Mfn2 ablation caused increased neuroinflammation in hippocampus and cortex in vivo. A Representative immunohistochemistry of anti-NeuN antibody in the hippocampus of Mfn2 cKO and control mice for immediate comparison with the neuroinflammation marker. b-e Representative immunocytochemistry of anti-GFAP antibody in the hippocampus (b) and cortex (d) and anti-Iba1 antibody in the hippocampus (c) and cortex (e) of the Mfn2 cKO mice and control mice

Figure 2.10 Quantification of the immunostaining of (A) GFAP, (B) IBA-1 and (C) MAP2 in the CA1 regions of the Mfn2 cKO mice at different ages (8-28 weeks) compared to control mice at 28 weeks of age (Data represent mean ±SEM, Student’s t-test, *P < 0.05). Chapter 3. DLP1 Cleavage by Calpain in AD

Figure 2.11 Mfn2 ablation caused abnormal cytoskeletal alterations in hippocampus and cortex in vivo. Representative immunohistochemistry of anti-MAP2 antibody in the hippocampus (a) and cortex (b) of Mfn2 cKO and control mice. Representative immunohistochemistry of anti-AT8, an antibody specifically against phosphorylated tau, in the cortex (c) of Mfn2 cKO and control mice

Chapter 3. DLP1 Cleavage by Calpain in Alzheimer’s Disease

Abstract

Abnormal mitochondrial dynamics contributes to mitochondrial dysfunction in

Alzheimer’s disease (AD), yet the underlying mechanism remains elusive. In the current

study, we reported that DLP1, the key mitochondrial fission GTPase, is a substrate of

calpain which produced specific N-terminal DLP1 cleavage fragments. In addition, various AD-related insults such as exposure to glutamate, soluble Aβ oligomers or reagents inducing tau hyperphosphorylation (i.e., okadaic acid) led to calpain-dependent

cleavage of DLP1 in primary cortical neurons. DLP1 cleavage fragments were found in

cortical neurons of CRND8 APP transgenic mice which can be inhibited by calpeptin, a

potent small molecule inhibitor of calpain. Importantly, these N-terminal DLP1 fragments were also present in the human brains and the levels of both full length and N- terminal fragments of DLP1 were significantly reduced in AD brains along with full length and calpain-specific cleavage product of spectrin. These results suggest that calpain-dependent cleavage is at least one of the posttranscriptional mechanisms that contribute to the dysregulation of mitochondrial dynamics in AD.

Background

Alzheimer’s disease (AD), the most common cause of dementia in the elderly, is characterized by progressive neurodegeneration and cognitive impairment. The pathological hallmarks of AD include accumulation of extracellular plaques composed of amyloid beta (Aβ) and intracellular neurofibrillary tangles (NFT) composed of hyperphosphorylated tau in the hippocampus and cortex of the human brain(Mattson

2004). There is currently no cure or viable treatment for the neurodegeneration and progressive dementia in AD. According to the World Alzheimer’s Report there is an ever-increasing number of people living with AD and will increase to 131.5 million by

2050 (Salthouse 2004).

Although the exact mechanism of AD remains elusive, much research has been done to implicate mitochondrial dysfunction as an early prominent feature in susceptible neurons that plays a critical role in the pathogenesis of the disease(Silva et al. 2012;

Swerdlow 2016; Wang et al. 2014). Such a primary role is underscored by the fact that defective glucose utilization and energy metabolism is a well-documented abnormality preceding functional impairment in patients with mild cognitive impairment, a prodromal stage of AD, and in AD(Swerdlow 2016). While mechanisms underlying mitochondrial dysfunction in AD remain incompletely understood, recent studies from multiple groups demonstrated that abnormal mitochondrial dynamics and distribution are likely involved: overexpression of familial AD APP mutations or exposure to soluble Aβ oligomers induced profound mitochondrial fragmentation, ultrastructural damage and reduced mitochondrial distribution in neuronal processes in neuronal culture (Wang et al. 2009;

Cho et al. 2009; Barsoum et al. 2006; Wang et al. 2008; Manczak et al. 2010; Du et al.

2010; Calkins & Reddy 2011). These deficits are causally involved in Aβ-induced mitochondrial dysfunction and synaptic abnormalities in primary hippocampal or cortical neurons in vitro (Wang et al. 2008). Mitochondrial damage in the form of irregular distribution and round and engorged mitochondria are also documented in AD mouse models(Trushina et al. 2012; Wang et al. 2017). Mitochondrial numbers were reduced and became dystrophic and fragmented in the vicinity of plaques in APP/PS1 mice revealed by real time imaging study(Xie et al. 2013). Similarly, swollen mitochondria with extensive ultrastructural damage and abnormal distribution are also observed in the brain of AD patients (Hirai et al. 2001; Xinglong Wang et al. 2009).

Mitochondria are dynamic organelles that undergo fusion and fission controlled by large GTPases: mitochondrial fission is regulated by cytosolic protein dynamin like protein 1 (DLP1) which translocates to mitochondrial outer membrane during fission with the assistance of mitochondrial outer membrane proteins such as Fis1 or

Mff1(Mishra & Chan 2014). Mitochondrial fusion is regulated by mitofusin 1 and 2

(MFN1/2) on the outer mitochondrial membrane and OPA1 on the inner mitochondrial membrane(Mishra & Chan 2014). Mitochondrial dynamics is critical for maintaining the homeostasis of mitochondria including the tight regulation of their morphology and distribution according to the metabolic need of the cells. Changes in mitochondrial dynamics significantly impact almost all aspects of mitochondrial function, and defects in the large GTPases involved in mitochondrial fission/fusion cause human neurological diseases(David C Chan 2006). Interestingly, our prior studies revealed that all these large

GTPase involved in mitochondrial fission and fusion are decreased in the brain of AD patients and in soluble Aβ oligomer-treated cells(Xinglong Wang et al. 2009). It appears

that these changes not only result in mitochondrial fragmentation, but also lead to

mitochondrial abnormal distribution since DLP1 overexpression could rescue Aβ-induced

depletion of mitochondria in the neuronal processes(Wang et al. 2008). Our recent study

demonstrated that Mfn2 ablation caused neurodegeneration and other AD-related deficits in the hippocampus and cortex (Jiang et al. 2018). However, how these large GTPases become reduced in AD is unknown.

Calcium dyshomeostasis has been well documented in AD as many of the clinical mutations in the presenilin (PS1/PS2) genes have been shown to disrupt the calcium signaling cascade(LaFerla 2002). The aberrant calcium signaling leads to excitotoxicity and may be a mechanism of neuronal death in AD(Mattson et al. 2000). It is well established that Aβ oligomers induce a rapid and sustained increase in intracellular calcium in neurons(Li et al. 2011; Kelly & Ferreira 2006) and Aβ induces neuronal abnormalities including spine loss and synaptic dysfunction likely through the activation of calcium-dependent signaling molecules such as calpain (Granic et al. 2010; Liang et al. 2010; Ferreira 2012). Interestingly, there is evidence that calpain activation leads to cleavage of the dynamin protein in an AD cell model (Kelly et al. 2005). Given the similarity between dynamin and dynamin-related proteins involved in mitochondrial fission and fusion, we hypothesized that calpain activation may be involved in the reduction of these mitochondrial fission/fusion proteins in AD. Therefore, in this study we aimed to investigate the posttranscriptional regulation of mitochondrial fission/fusion

GTPases with a focus on the effects of calpain activation and reduced levels of DLP1, the key protein involved in mitochondrial fission and distribution. We observed the cleavage of DLP1 by calpain into several fragments at ~65 kDa and 50 kDa in size. This cleavage

75 can be triggered by AD-relevant insults such as exposure to glutamate, soluble Aβ oligomers or agents inducing tau phosphorylation. Cleavage fragments of DLP1 as well as the cytoskeletal protein, spectrin, are observed in AD patient brains supporting our theory that calpain is a protease that plays a role in the reduction of mitochondrial proteins seen in AD.

Methods

Human brain tissues

Samples were obtained at autopsy and following approved IRB protocols, were received from the Brain Bank at Case Western Reserve University and from the NIH

Neurobiobank in accordance with the institutional bioethics guidelines. The diagnosis of

Alzheimer’s disease was obtained according the NINCDS-ADRDA group criteria(McKhann et al. 1984). Samples of frozen cortical gray matter of AD and age- and gender-matched control cases (n = 7/group) were homogenized and lysed with RIPA

Buffer (Cell Signaling) plus 1 mM phenylmethylsulfonyl fluoride (Sigma) and Protease

Inhibitor Cocktail (Sigma) and centrifuged for 10 min at 16,000 × g at 4 °C. Protein concentrations of the lysates from total cortical gray matter homogenates were determined by the bicinchoninic acid assay method (Pierce, Rockford, IL, USA). Equal amounts of proteins (20 μg) were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to immobilon membranes. After blocking with 10% non-fat dry milk, primary and secondary antibodies were applied and the blots developed with enhanced chemiluminescence (Santa Cruz, Dallas, Texas, USA).

Primary rat and CRND8 mouse neuronal culture

Primary cultures of both rat and mouse cortical neurons were prepared from the brains of embryonic pups at day 18 and day 16, respectively, as previously reported

(Kaech & Banker 2006; Zhao et al. 2017) with some modification. In brief, the cerebral cortices were dissected from the embryonic brain and dissociated by trypsinization for 10 min at room temperature. The resulting cell suspensions were resuspended in neurobasal

77 medium supplemented with B27 (Gibco-BRL, Waltham, Massachusetts, USA) and penicillin-streptomycin (Thermo Fisher Scientific, Waltham, MA, USA)) and plated onto poly-d-lysine (Sigma, St. Louis, MO, USA) coated plastic plates. Neurons were maintained at 37 °C in 5% CO2 for 12 days prior to chemical treatment. After 12 days, the neuron culture medium was replaced with fresh medium containing treatments of Aβ, glutamate, and calpeptin as described in the results and figure legends.

Drugs

Glutamate (50 μM; Sigma, St. Louis, MO, USA), calpeptin (50 μM; Tocris

Bioscience, Minneapolis, MN, USA), and okadaic acid (5-50nM; Cayman Chemical,

Ann Arbor, MI, USA) were added to neuronal cultures at the indicated final concentrations. Aβ oligomers were prepared as previously described(Song et al. 2014).

Briefly, lyophilized Aβ peptides were dissolved in dimethyl sulfoxide, diluted in neurobasal without phenol red (Gibco-BRL, Waltham, Massachusetts, USA) to a final concentration of 1 μm and incubated at 4 °C for 16 h. Prepared Aβ oligomers were added to neuronal cultures for the indicated times.

Calpain cleavage assay

In vitro cleavage of recombinant GST-DLP1 protein by calpain was performed as previously described(Garg et al. 2011). Briefly, recombinant GST-DLP1 (160 ng;

Abnova, Walnut, CA, USA) was incubated with calpain-1 (Biovision, San Francisco,

CA, USA) in reaction buffer for various times with or without calpeptin (50μm; Tocris

Bioscience, Minneapolis, MN, USA). After being incubated for the indicated times, the

reaction mixture was mixed with an equal volume of 2 × SDS sample buffer and boiled

for 10 min. Samples were subjected to SDS-PAGE followed by Western blotting with

anti-DLP1 D6C7 (Cell Signaling Technology, Danvers, MA, USA) or anti-GST (Santa

Cruz, Dallas, Texas, USA) antibodies.

Western blotting

Cell lysates from primary neurons were prepared with protein extraction solution

(Cell Signaling Technology, Danvers, MA, USA) in accordance with the manufacturer’s

guidelines. Proteins were subjected to SDS-PAGE and subsequently transferred to PVDF membrane (Bio-Rad, Hercules, CA, USA) and blocked with 5% skim milk in TBST buffer. Blots were incubated for 16 h at 4 °C with primary antibodies to DLP1 D6C7

(1:1000; Cell Signaling, Danvers, MA, USA), Spectrin (1:1000; Cell Signaling), Actin

C4 (1:5000; Thermo Fisher Scientific, Waltham, MA, USA), DLP1 C-5 (1:1000; Santa

Cruz, Dallas, Texas, USA). The blots were washed in TBST buffer, incubated with secondary antibodies for 1 h at 23 °C and visualized using enhanced chemiluminescence reagents (Santa Cruz, Dallas, Texas, USA).

Statistical analysis

Data are presented as means ± standard error of the mean (SEM) of at least three independent experiments and, where appropriate, were analyzed using one-way analysis of variance (ANOVA) followed by Student's t-test. P < 0.05 was considered statistically significant.

Results

Cleavage of recombinant DLP1 by Calpain

To investigate whether DLP1 is directly cleaved by calpain, we incubated

recombinant N-terminal GST-tagged DLP1 with calpain-1 in reaction buffer for various

times and concentrations and examined the DLP1 levels by Western blot analysis. Using

a C-terminal DLP1 antibody (i.e., DLP1 C5 antibody against amino acids 560-736 region

of DLP1) a significant reduction in the level of full length GST-DLP1 is observed with

0.05 units (0.05U) of calpain-1 after a 30-minute treatment, which became more

significantly reduced with higher units of calpain-1(Figure 3.1A). In fact, 30-minute

treatment with 0.25 or 0.5 units of calpain-1 resulted in total loss of full length GST-

DLP1, reflecting a dose-dependent effect. We also used a GST antibody to detect GST-

DLP1 after cleavage by calpain. Interestingly, in addition to the dose-dependent

reduction in the level of full length GST-DLP1, we also observed the appearance of a

specific band around 75kDa with the 30 min treatment of 0.05 units of calpain-1 which peaked with the treatment of 0.25 units of calpain-1 but then decreased with the treatment of 0.5 units of calpain-1. Since the DLP1 was N-terminally tagged with GST, this 75 kDa band likely reflects a 50 kDa N-terminal cleavage fragment of DLP1. The reduction of full length GST-DLP1 and appearance of cleavage fragments were completely prevented when calpeptin, a calpain inhibitor, was present along with calpain, demonstrating the specificity of the calpain cleavage reaction. Treatment with 0.25 units of calpain-1 led to gradual reduction of full length GST-DLP1 with time until its full cleavage after 10 minutes as revealed by the C-terminal DLP1 antibody (Figure 3.1B). Similarly, we also

observed the appearance of the 75 kDa fragment as early as 1 minute, which peaked at 10

minutes and was completely gone at 15 minutes (Figure 1B).

To further investigate this calpain-dependent cleavage of DLP1, whole cell lysate from M17 neuroblastoma cells was incubated with calpain-1 at varying concentrations for 30 minutes. Using an antibody against N-terminal DLP1 (i.e., DLP1 D6C7 antibody), we confirmed the dose dependent reduction of full –length DLP1 by calpain-1 (Figure

3.2). Importantly, accompanying the reduced full length DLP1, there was an accumulation of a major 50 kDa band after calpain-1 treatment. In fact, the accumulation of this band even persisted with up to 4 hours of 1 unit calpain-1 treatment (not shown).

There was also a weak band around 65 kDa in the lysate treated with 0.2 units calpain which was gone in the lysate treated with 0.5 units calpain, likely reflecting a transient intermediate DLP1 fragment generated by calpain cleavage. In fact, these two cleavage products were also present in the M17 lysate at basal condition as faint bands, suggesting such cleavage likely occurs endogenously. The concurrent treatment with calpeptin abolished both the reduction of full length DLP1 and the appearance/accumulation of

DLP1 N-terminal fragments. These results suggest that DLP1 is a substrate of calpain and that calpain cleavage of DLP1 yields N-terminal fragments of 65 and 50-kDa in size. We therefore used DLP1 N-terminal antibody to detect these calpain-dependent N-terminal

DLP1 fragments in later studies.

DLP1 cleavage by Calpain in Glutamate treated neurons

To investigate calpain-dependent DLP1 cleavage in cells, we examined the DLP1 protein levels in primary neurons treated with glutamate, which induces calcium

dyshomeostasis and excitotoxicity through the activation of the NMDA and AMPA

receptors. Under this condition, glutamate treatment led to activation of calpain activity

as revealed by the almost complete loss of 250 kDa full length spectrin protein and the

appearance of a major 150 kDa spectrin cleavage product which has been reported to be

specific to calpain activity(Rajgopal & Vemuri 2002) (Figure 3.3A,B). Similarly, the full

length DLP1 protein was almost completely gone in glutamate-treated cells which was

accompanied by an accumulation of the 50 kDa N-terminal fragment as was identified in

both the cell free assay and M17 cell lysate (Figure 3.2). With the addition of calpeptin to

the glutamate treatments, the loss of full length spectrin and DLP1 along with the

increase in the specific cleavage fragments of these two proteins were significantly

inhibited (Figure 3.3A-C).

DLP1 cleavage by Calpain in Aβ- and okadaic acid- treated neurons

To investigate a potential mechanism underlying DLP1 reduction in AD, we

examined DLP1 levels in the primary neurons treated with soluble Aβ oligomers. As

expected, treatment with Aβ oligomers led to decreased full length spectrin and a

significant increase in the 150 kDa spectrin cleavage fragment (Figure 3.4A,B) indicating

that treatment with Aβ caused calpain activation. Similarly, a significant increase in the

50-kDa cleavage fragment was revealed when detected by the DLP1 N-terminal antibody

(Fig 3.4C). Co-treatment with calpeptin in Aβ-treated neurons prevented the activation of

calpain as demonstrated by the restoration of full length spectrin along with the reduction

of 150 kDa cleavage fragment to the level similar to vehicle-treated cells. Under this condition, the appearance of DLP1 cleavage fragment was also completely reversed.

Hyperphosphorylation of tau protein is also involved in the pathogenesis of AD.

We further analyzed the impact of phosphorylated tau on calpain cleavage of DLP1. To

mimic the neurotoxicity of phosphorylated tau in cell cultures, we treated primary rat

neurons with okadaic acid which is inhibitor of serine/threonine phosphatases 1 and 2A

and induces hyperphosphorylation of tau(Kamat et al. 2013). Interestingly, okadaic acid

treatment also induced the degradation of both full length spectrin and DLP1 proteins in

cultured neurons (Fig 3.5A,B) and the appearance of calpain cleavage products. As a

specificity control, concurrent calpepetin treatment abolished the cleavage of both

spectrin and DLP1 after okadaic acid treatment.

Calpain Activation in neurons of APP Transgenic Mouse Model

We next investigated whether the calpain-dependent cleavage of DLP1 is present

in neurons of CRND8 mice, a widely used APP transgenic (Tg) mouse model expressing

the APP Swedish (KM670/671NL) and APP Indiana (V717F) mutations that shows early

signs of memory impairment as well as striking amyloid plaque pathology(Chishti et al.

2001). In the primary cortical neurons isolated from CRND8 mice, we observed

significant presence of the 150-kDa calpain-dependent cleavage fragment of spectrin compared to neurons from their non-Tg littermate controls (Fig 3.6A,B) suggesting a higher calpain activity in the CRND8 neurons. In conjunction with activated calpain, we also observed a significant presence of the 65 and 50-kDa cleavage fragments of DLP1

(Fig 3.6A,C). Furthermore, treatment with calpeptin inhibited the increased levels of 150 kDa cleavage fragment of spectrin as well as the 65- and 50kDa DLP1 cleavage fragments (Fig 3.6A-C).

DLP1 Cleavage Fragments in AD Brains

Our prior studies demonstrated that proteins levels of DLP1 are significantly

reduced but mRNA levels of DLP1 are not changed in the brain of human AD patients

compared to age-matched control patients, suggesting the involvement of post-

transcriptional regulation of DLP1 expression. To investigate whether calpain activation

is involved in the reduced expression of DLP1 in AD brain, we performed western blot

analysis of spectrin and DLP1 using brain homogenates. Both the full length and calpain-

dependent 150 kDa cleavage fragments were present in both AD and control samples.

There was a trend (p=0.07) of decrease in the level of full length spectrin and a

significant decrease in the level of 150-kDa cleavage product in the brain tissues from

AD patients as compared to that from the age-matched non-AD control patients (Fig

3.7A-C). Similarly, our N-terminal DLP1 antibody revealed that both full length and the two cleavage fragments (i.e., 65 and 50 kDa bands) of DLP1 were also present in both

AD and control sample. Consistent with our previous report, there was a significant decrease in the level of full length DLP1. Moreover, the levels of both the 65 and 50-kDa cleavage fragments of DLP1 decreased in the AD samples as compared to the controls samples (Fig 3.7A,D,E).

Discussion

In this study, we investigated the association between DLP1 reduction and calpain

activation in the context of AD models and AD brain. By employing the in vitro calpain

cleavage assay using GST-DLP1 recombinant protein and cell lysate from M17 neuroblastoma cells, we demonstrated that DLP1 is indeed a substrate of calpain which produced specific N-terminal DLP1 fragments. Furthermore, various AD-related insults such as exposure to glutamate, soluble Aβ oligomers or reagents inducing tau hyperphosphorylation (i.e., okadaic acid) led to DLP1 cleavage accompanied by calpain activation in primary cortical neurons. DLP1 cleavage fragments were also present in cortical neurons of CRND8 APP transgenic mice. Importantly, calpeptin, a potent small molecule inhibitor of calpain, prevented DLP1 cleavage in all these models, demonstrating the specific involvement of calpain activation in DLP1 cleavage in these models. Lastly, we found the presence of N-terminal DLP1 fragments in the human brains and the levels of both full length and N-terminal fragments of DLP1 were significantly reduced in AD brains. These results suggest that calpain activation is at least one of the posttranscriptional mechanisms that contribute to the dysregulation of mitochondrial dynamics in AD.

The major finding in this study is that we firmly established DLP1 as a physiological and AD-relevant pathophysiological substrate of calpain in cells and in the brain. This was first clearly shown by our in vitro calpain cleavage assay demonstrating a dose- and time-dependent cleavage of recombinant GST-DLP1 by calpain which was inhibited by calpeptin. This in vitro assay also suggested the presence of an intermediate

50 kDa N-terminal DLP1 cleavage fragment before its complete digestion by calpain.

Indeed, we found this 50 kDa fragment along with another 65 kDa N-terminal fragment of DLP1 in the M17 cells at basal condition which could be enhanced by calpain treatment and inhibited by calpeptin, suggesting that they are present in the physiological condition by calpain cleavage. These two specific N-terminal fragments of DLP1 were also present in the brains from elderly human control patients which confirms the physiological presence of these cleavages in vivo. Importantly, the current study demonstrated that these calpeptin-sensitive cleavages of DLP1 were significantly enhanced by AD-relevant insults such as treatment with glutamate, soluble Aβ oligomers or okadaic acids, modeling excitotoxicity, Aβ exposure or tau hyperphosphorylation induction, which thus demonstrated the involvement of the calpain-dependent cleavage of

DLP1 in AD models. The pathophysiological relevance of calpain-dependent cleavage of

DLP1 is further corroborated by the evidence from cortical neurons of CRND8 APP transgenic mice and brain tissues from AD patients. Calpains have been closely linked with AD as they have been shown to cleave: APP that regulates Aβ production(Morales-

Corraliza et al. 2012), tau leading to production of neurotoxic fragments(Ferreira & Bigio

2011), synaptic proteins like dynamin-1(Kelly et al. 2005), and NMDA receptor subunit

NR2B important for synaptic health(Simpkins et al. 2003). Our results thus added reduced DLP1 as a victim of enhanced calpain activation in AD which affect mitochondrial dynamics and contribute to mitochondrial dysfunction during AD pathogenesis.

Previous studies of post mortem AD patient brains show increased calpain activity in end-stage AD brain(Atherton et al. 2014; Jin et al. 2015; Saito et al. 1993) as well as elevated levels of calpain substrates that are cleaved(Atherton et al. 2014; Liang

et al. 2010; F. Liu et al. 2005). In the current study, while we observed clear reduction, sometimes a complete depletion, in the levels of full length spectrin or DLP1 in the brains from AD patients, we also found significantly reduced levels of calpain-specific 150 kDa

fragment of spectrin along with reduced levels of two N-terminal fragments of DLP1.

This complete depletion of DLP1 full length and N-terminal fragments in nearly all of the

AD samples we examined, in our opinion, implicated excessive calpain activation that resulted in more complete degradation of DLP1 in AD. Along this reasoning, it was demonstrated that calpain activation is associated with disease progression through various Braak stages and the levels of full length and calpain-dependent cleavage products of certain substrates such as CAST peaked at earlier stages but both declined at later stage (Kamat et al. 2013). It is therefore of interest to further determine how the levels of DLP1 and its cleavage fragments change along different stages during the course of AD. Nevertheless, our study does not preclude the potential involvement of other proteases in the cleavage of DLP1. For example, studies have shown that caspases may indirectly regulate cleavage of mitochondrial GTPases such as OPA1(Loucks et al.

2009) and DLP1 during apoptosis(Estaquier & Arnoult 2007).

The presence of 65 and 50 kDa DLP1 fragments at physiological conditions, especially in the brain of elderly controls, may be interesting. The function of DLP1 has been thoroughly explored in both yeast and mammals as it is similar to the classic dynamins in membrane scission and remodeling of mitochondria(Sesaki et al. 2014). It is

a cytosolic protein that forms dimers and tetramers and is recruited to the mitochondrial

surface through interactions with various outer mitochondrial proteins such as

mitochondrial fission factor (Mff), Fis1, and mitochondrial elongation factor

(MIEF1/2)(Sesaki et al. 2014). Crystal structures show that DLP1 assembles into spirals which mediates mitochondrial scission(Mears et al. 2011). DLP1 contains four domains: an N-terminal GTPase domain, middle domain, variable domain, and C-terminal GTPase effector domain (GED)(Elgass et al. 2013; Otera et al. 2013). The middle domain is important in the regulation of DLP1 self-assembly into dimers and tetramers(Ford et al.

2011; Fröhlich et al. 2013; Faelber et al. 2011). So it can stand to reason that cleavage of

DLP1 somewhere in the middle domain resulting in 50kDa and 65kDa fragments as shown in our data could lead to dysfunctional oligomerization of DLP1 and therefore affect its function in regulating mitochondrial fission. While DLP1 knockdown could lead to abnormal mitochondrial distribution, how DLP1 cleavage fragments may affect mitochondrial transport and distribution is not clear. Further studies will be needed to map the specific cleavage sites and their impact on DLP1 functions.

In conclusion, this study established DLP1 as a calpain substrate and suggested that calpain activation could contribute to reduced DLP1 levels in AD and contribute to mitochondrial dynamics abnormalities.

Figure 3.1 Dose- and time-dependent cleavage of recombinant DLP1 by calpain-1. (A) 160 ng recombinant GST-DLP1 was incubated with calpain-1 at 30 °C at the indicated concentration for 30 minutes in the absence/presence of calpeptin and the remaining GST-DLP1 was visualized by western blotting with a C-terminal specific DLP1 antibody (upper panel) or a GST antibody (lower panel). (B) 160 ng recombinant GST-DLP1 was incubated with 0.25 units calpain-1 at 30 °C for the indicated duration and the remaining GST-DLP1 was visualized by western blotting with a C-terminal specific DLP1 antibody (upper panel) or a GST antibody (lower panel).

Figure 3.2 DLP1 is cleaved by calpain in M17 neuroblastoma cell lysates after incubation with calpain-1. Representative immunoblot of DLP1 N-terminal specific antibody (i.e., DLP1 D6C7) in whole cell lysates from M17 cells incubated with calpain- 1 at 30 °C at the indicated concentrations for 30 min in the absence/presence of calpeptin.

Figure 3.3 Calpain-dependent cleavage of spectrin and DLP1 in glutamate-treated rat primary cortical neurons. (A) Representative western blots of DLP1 in primary cortical neurons (DIV12) treated with 50 µM glutamate in the absence/presence of 50 µM calpeptin for 4 h. Spectrin was probed as a positive control for calpain activation; β-tubulin was probed as an internal loading control. (B-C) Quantitative analysis of levels of 150 kDa cleavage fragment (CF) of Spectrin CF (B) and 50 kDa cleavage fragment (CF) of DLP1 (C). The expression levels of all proteins were normalized to tubulin and expressed as a relative to non-treated control values of each respective protein. Data are presented as the mean ± SEM of three independent experiments (*p < 0.05, **p < 0.001).

Figure 3.4 Calpain-dependent cleavage of spectrin and DLP1 in rat primary cortical neurons treated with soluble Aβ oligomers. A) Representative Western blots of DLP1 in primary cortical neurons (DIV12) treated with 1 µM soluble Aβ oligomers in the absence/presence of 50 µM calpeptin for 24hr. Spectrin was probed as a positive control for calpain activation; Actin was probed as an internal loading control. FL, full-length DLP1 or Spectrin; CF, cleavage fragment of DLP1 or Spectrin. (B-C) Quantitative analysis of levels of Spectrin CF (B) and DLP1 CF (C). The protein levels were normalized to actin and non-treated control values of each respective protein (B-E). Data are presented as the mean ± SEM (*P < 0.05, **P < 0.001).

Figure 3.5 Calpain-dependent cleavage of spectrin and DLP1 in rat primary cortical neurons treated with okadaic acid. A) Representative Western blots of DLP1 in primary cortical neurons (DIV12) treated with okadaic acids at the indicated concentration in the absence/presence of 50 µM calpeptin for 24hr. Spectrin was probed as a positive control for calpain activation; actin was probed as an internal loading control. FL, full-length DLP1 or Spectrin; CF, cleavage fragment of DLP1 or Spectrin. (B) Quantitative analysis of levels of DLP1 CF. The protein levels were normalized to actin and non-treated control values of each respective protein. Data are presented as the mean ± SEM (*P < 0.05, **P < 0.001).

Figure 3.6 Calpain-dependent cleavage of spectrin and DLP1 in primary cortical neurons isolated from CRND8 APP transgenic mice. (A) Representative Western blots of DLP1 in primary cortical neurons (DIV12) isolated from either CRND8 mice or littermate control mice treated with/without 50 µM calpeptin for 24hr. Spectrin was probed as a positive control for calpain activation; actin was probed as an internal loading control. FL, full-length DLP1 or Spectrin; CF, cleavage fragment of DLP1 or Spectrin. (B-C) Quantitative analysis of levels of Spectrin CF (B) and DLP1 CF (C). The protein levels were normalized to actin and non-treated control values of each respective protein (B-C). Data are presented as the mean ± SEM (*P < 0.05, **P < 0.001).

Figure 3.7 Decreased level of DLP1 in Alzheimer’s Disease (AD) brain. (A) Expression of DLP1 and Spectrin in AD and age-matched control brains were analyzed by Western blotting with antibodies to DLP1 and spectrin. (B-E) Representative graphs showing quantification of levels of DLP1 FL (B), DLP1 CF (C), Spectrin FL (D), and Spectrin CF (E). Data are presented as the mean ± SEM (*P < 0.05, **P < 0.001).

Chapter 4. Summary, Future Directions, Conclusion

Mitochondrial Fragmentation Causes Neurodegeneration and other AD-Related

Pathologies in the Hippocampus and Cortex

It is well established that mitochondrial dysfunction is an early and prominent feature of AD. Indeed, abnormalities of almost all aspects of mitochondrial structure and function have been documented in AD models and brain tissues of AD patients which includes bioenergetics defects such as reduced activity and/or expression of rate-limiting enzymes in the TCA cycle and oxidative phosphorylation, extensive oxidative stress involving oxidative modifications to all types of bio-molecules and reduced

activity/expression of antioxidant enzymes and antioxidants, mitochondrial calcium

abnormalities, and increased prevalence of mtDNA mutations. Studies in the most recent

years focused on changes in mitochondrial dynamics and its potential contribution to mitochondrial dysfunction and the pathogenesis of AD. As we reviewed in Chapter 1, our group and others have demonstrated disruption of the mitochondrial fission/fusion balance in AD. We have previously shown that the abnormal mitochondrial distribution and morphology consistent with mitochondrial fragmentation are prevalent in AD patient brains (X. Wang et al. 2009). It has also been shown by our lab and several other groups that mitochondrial dynamic GTPases Mfn2, Mfn1, OPA1, and DLP1 expression levels are also changed in AD (Reddy et al. 2012; Bossy et al. 2010; X. Wang et al. 2009). In fact,

AD-related proteins such as Aβ and phosphorylated tau proteins directly interact with mitochondrial fission factors. However, given the snapshot nature of postmortem tissues used in these studies, it remains to be determined of the role that mitochondrial dynamics plays in the pathogenesis of AD as to whether it is the cause, the consequence, or a

bystander of the biochemical and morphological changes seen in AD.

A critical and causal role of mitochondrial dynamic changes in mitochondrial

dysfunction is clearly demonstrated in in vitro AD models: in neuroblastoma cells and in primary hippocampal or cortical neurons, the expression of fAD-causing APP mutants or exposure to soluble Aβ oligomers causes mitochondrial fragmentation, which is likely a combined consequence of excessive fission and reduced fusion. It is also clearly demonstrated in these in vitro models that mitochondrial fragmentation causes mitochondrial dysfunction such as reduced mitochondrial membrane potential and ATP production and increased mitochondrial oxidative stress since rescue of mitochondrial fragmentation by overexpressing OPA1 or treatment with mitochondrial division inhibitor

(i.e., mdivi-1) could alleviate these functional deficits (Wang et al. 2008). In vivo studies

also confirmed a tipped balance of mitochondrial fission and fusion in various APP

transgenic mouse models of AD: mitochondrial morphological deficits consistent with a

fragmented phenotype were reported in CRND8 mice as early as 3 months of age, well

before the development of amyloid plaques. A peculiar “beads-on-the-string” morphology of mitochondria was reported in APP/PS1 and Tg2576 mice (Trushina et al. 2012) which is likely due to an enhanced fission but stalled at the last step of outer membrane separation.

In vivo imaging also confirmed fragmented mitochondria in the vicinity of amyloid plaques in the APP/PS1 mice. More recently, studies from two different groups including ours demonstrated that treatment of mdivi-1, the mitochondrial division inhibitor, effectively alleviated the mitochondrial morphological deficits and amyloid pathology along with cognitive/behavioral deficits in two different APP transgenic mouse models (Baek et al.

2017; Wang et al. 2017). However, due to the lack of neuronal loss in these APP transgenic mouse models, it remains to be determined whether indeed mitochondrial dynamic

abnormalities lead to neurodegeneration in AD-affected brain areas. Aiming at this important gap in our knowledge, we hypothesized that altered mitochondrial morphology

(mitochondrial fragmentation) and distribution (reduced dendritic densities) are prominent features that could lead to the neurodegeneration in the hippocampus and cortex as seen in

AD.

To test this hypothesis, we tried to generate a mouse model with both mitochondrial fragmentation and abnormal distribution as we found in AD in the brain areas afflicted by the disease. Given the consistent results of reduced Mfn2 in AD among different groups, and the prior studies showing that knockout of Mfn2 in Purkinje neurons as well as dopaminergic neurons causes mitochondrial ultrastructural deficits (fragmentation) and abnormal distribution, and loss of bioenergetics (Pham et al. 2012; Chen et al. 2007), we

chose to specifically knockout Mfn2 from the hippocampus and forebrain by crossing the

CaMKII-alpha-cre mice with Floxed Mfn2 mice. Using electron microscopy, we were able

to observe that mitochondria in pyramidal neurons of the cKO mice had significantly larger

diameter and appear round as compared to the regular elongated and tubular mitochondria

in the control mouse neurons. This in fact resembles the slightly shorter but fatter (i.e.,

increased size) mitochondria observed in the pyramidal neurons of AD brain, reflecting a

swollen phenotype. Along those lines we also saw broken cristae in the inner structures of

these mitochondria in the pyramidal neurons in cKO mice with multilamellar appearance

and vacuolation, again resembling that of AD neurons. Importantly, in these cKO mice we

saw selective and severe neurodegeneration starting at 12 weeks of age and at 18 weeks

almost complete loss of CA1 hippocampal neurons. Preceding this neuronal loss, however,

we see a progression starting with abnormal mitochondrial dynamics where we begin to

99 see loss of mitochondria in the neuronal axons at 8 weeks. At 8 weeks we begin to see a loss of mitochondrial bioenergetics using both Western blotting analysis as well as an extracellular flux analyzer to measure oxygen consumption. Starting at 8 weeks we also observe an increased level of oxidative stress as measured by immunohistochemistry staining as well as using a protein oxidation detection system. Before the neurodegeneration at 18 weeks, we also observe infiltration of inflammatory cells starting as early as 8 weeks with GFAP-positive staining for gliosis becoming significant at 12 weeks in the hippocampus and 18 weeks in the frontal cortex. IBA1-positive stained activated microglial cells appear at 12 weeks and becomes profuse at 18 weeks. These IHC results show us the progression towards neurodegeneration after mitochondrial dysfunction.

Finally using IHC we again show as early as 12 weeks there is a microtubule deficit seen by the abnormal distribution of MAP2 protein suggesting cytoskeletal changes precede cell death. These results thus demonstrated a strong causal relationship between disruption of mitochondrial fission/fusion and neurodegeneration in vivo in the hippocampus and cortex.

Further, we also establish a timeline of events: starting from mitochondrial morphological and distribution changes, to disrupted to bioenergetics, to oxidative stress, to inflammatory infiltrates, to cytoskeletal changes, and finally to neurodegeneration. It must be emphasized that all these events also occur in the course of AD.

While our study clearly established the causality of neurodegeneration caused by

Mfn2 knockout in AD-afflicted brain areas such as hippocampus and cortex, there are several important caveats: firstly, deficits observed in the brain of Mfn2 cKO mice are due to the total loss of Mfn2 in affected neurons. However, while Mfn2 level is reduced in AD brain, it is unlikely that Mfn2 is completely depleted in pyramidal neurons in the brain of

100

AD patients. Given that hemizygous Mfn2 cKO mice demonstrated no neurodegeneration

even at 13 months of age (data not shown), neurodegeneration in AD is unlikely only due

to Mfn2 reduction. Mitochondrial dynamic abnormalities and abnormal distribution are the

consequence of combined effects of altered expression of multiple fission/fusion factors

which could not be faithfully modeled in vivo. Secondly, Mfn2 cKO mice have Mfn2

depleted in affected neurons very early in life (i.e., after CAMKII expression around 4-6

weeks) while AD neurons experience reduced Mfn2 or altered fission/fusion proteins much

later in life. Thirdly, there are concerns on the multiple functionality of Mfn2. One of the

major functions is mediating mitochondrial fusion, but how are the other functions of Mfn2

affected and how do those functions play a role in neurodegeneration? Mfn2 has been

shown to be involved in mitochondrial trafficking as it interacts with the Miro/Milton

complex (Misko et al. 2010). If Mfn2 is knocked out, we would assume there would be a down regulation of mitochondrial trafficking, which is something that we in fact observe in the neurons. The abnormal distribution of mitochondria could indeed be a result of the loss of Mfn2 in the transport complex. Another role that Mfn2 presumably plays, although controversially, is to act as a tether between mitochondria and the endoplasmic reticulum

(ER). It has been debated whether loss of Mfn2 increases or decreases mitochondria-ER

(MAM) contacts (de Brito & Scorrano 2008; Leal et al. 2016), but our knockout mouse

could definitely help answer that question. Using electron microscopy, we can quantify

MAM structures and using Western blotting look at levels of specific MAM proteins. This

will definitely help in answering the question of whether or not Mfn2 is a positive or

negative regulator of MAM. However, whatever the changes in MAM are in Mfn2 cKO

mice, it will complicate the interpretation of the results. Another role that Mfn2 plays is

101

that of a mitophagy regulator: either by ubiquitination by PINK1/parkin leading to

disruptions of MAM and increasing mitophagy (McLelland et al. 2018) or recruiting Parkin

to damaged mitochondria thus promoting the lysosomal degradation of these mitochondria

(Pallanck 2013). In our knockout model, we would expect that loss of Mfn2 leads to a decrease in mitophagy because of the factors described above. Loss of mitophagy would result in an increased accumulation of damaged mitochondria which is something that we observe, so we could definitely start by looking at the levels of ubiquitinated levels of Mfn2 to see if they are indeed decreased in our knockout. We could also perform imaging studies and look to see whether there is a decrease in formation of autophagosomes. The loss of mitophagy due to a loss of Mfn2 would definitely be a good explanation for the accumulation of damaged mitochondria that we see in our EM pictures. Finally, another experiment that could be done to try and prove causality would be to overexpress Mfn2 or even Mfn1 in the hippocampal or cortical neurons to see whether or not that would rescue the neurodegeneration phenotype. It has been shown that Mfn1 overexpression can indeed rescue Mfn2 knockout phenotypes in Purkinje neurons (Chen et al. 2007), therefore we expect that Mfn1 would also be sufficient in rescuing the neurodegeneration phenotype in our knockout mouse.

102

Initial Exploration of Mechanisms Underlying Abeta-Induced Mitochondrial

Dynamics Abnormalities: Focusing on Calcium Signaling and Calpain Activation

Mitochondrial dynamic abnormalities are well documented in the in vitro and in vivo models of AD as well as in the brain of AD patients. In Chapter 3, we established that a complete loss of Mfn2 in the pyramidal neurons leads to neurodegeneration likely through mitochondrial fragmentation and abnormal distribution. We recently also demonstrated that the inhibition of mitochondrial fragmentation by mdivi-1 treatment could prevent AD-related deficits such as pathological changes and cognitive/behavioral deficits in CRND8 mice. Given the critical role of mitochondrial dynamics in AD pathogenesis, it is of paramount importance to understand the mechanism underlying mitochondrial dynamic abnormalities in AD. Indeed, our studies demonstrated that all the large GTPases involved in mitochondrial fission/fusion are reduced in AD which could cause abnormal mitochondrial distribution and contribute to fragmentation phenotype in combination. However, how these proteins become dysregulated in AD has not been studied.

Calcium dyshomeostasis is a very well documented finding in AD (LaFerla 2002) due to the mutations of presenilin 1 and 2 causing a destabilization of the phosphoinositide signaling cascade. Many of the early studies on calcium dyshomeostasis centered around the mutations of PS1/PS2 on fAD which lead to increased intracellular calcium levels through ER release of calcium (Chan et al. 2000; Cedazo-Minguez et al. 2002). This

calcium hypothesis of AD suggests that these changes in calcium may lead to

consequences in plasticity as well as long-term potentiation. In vivo studies using a triple

103

transgenic mouse expressing APP, tau, and PS1 mutants show that there is indeed a

significantly higher level of intracellular rise of calcium that may be related to an

increased level of ryanodine receptor (RyR) on the ER surface (Smith et al. 2005). Based on these studies, a key question begins to arise as to what pathophysiological consequence comes from this increased calcium and how does that affect the onset or progression of AD. This is where Aβ appears to provide the link between calcium dyshomeostasis and AD progression in sporadic cases. Studies have shown that Aβ can act upon important calcium channels such as the voltage-gated calcium channel (Ueda et al. 2002), NMDA receptors (Wu et al. 1995), and nicotinic receptors (Dougherty et al.

2003) leading to an increased calcium influx. Within the cell, Aβ has been shown to cause an increased release of calcium via its interaction with IP3R in SH-SY5Y cells

(Smith et al. 2001). Having established a connection between Aβ activity and an increased level intracellular calcium, the next question is to explore the downstream consequences of calcium signaling. There are both pathogenic and protective signaling cascades following calcium influx. First the cell tries to compensate by activating anti- apoptotic factors such as NF-κB (Pahl & Baeuerle 1996), Bcl-2, and cAMP response element-binding protein (CREB) (Mattson & Furukawa 1997). The pathogenic responses include expression of C/EBP homologous protein (CHOP) that inhibits Bcl-2 and other protective proteins and increases mitochondrial calcium overload leading to generation of

ROS and apoptotic signals such as caspases and cytochrome c (Stutzmann 2007). Many of these studies correlate Aβ toxicity with an increased calcium influx, but there is a knowledge gap of how calcium signaling leads to some of the other adverse effects of Aβ toxicity such as mitochondrial dysfunction in AD.

104

In Chapter 4, we hypothesized that the loss of the large mitochondrial GTPases

that regulate fission/fusion may be due to a calcium-activated process. It has been shown

previously in primary neurons as well as a transgenic mouse that Aβ-induced activation of calpain leads to its cleavage of dynamin 1 (Kelly et al. 2005). Therefore, we believe that the dynamin-related GTPases such as Mfn2 and DLP1 could also be cleaved by calpain leading to their reduction. Our lab has shown that Mfn2 can be cleaved by calpain by a glutamate induced calcium influx in mouse motor neurons (Wang et al. 2015). So it would be not a stretch to believe that Aβ-induced calcium influx could also activate calpain. To test this hypothesis, we used biochemical techniques to show that calpain is indeed activated and that these GTPases are substrates that can be cleaved by calpain. By using a cell-free assay, we were able to show that purified GST-tagged DLP1 is indeed cleaved by purified calpain 1 and that cleavage produces very prominent and specific cleavage products at 50kDa and 60kDa. Using the appearance of these cleavage products as well as cleaved spectrin as an indicator for calpain activation, we next went on to show that this cleavage of DLP1 and Mfn2 by calpain can be observed in vitro in M17 neuroblastoma cell lines. Using glutamate-induced calcium influx as a control, we showed in primary neurons that the activation of calpain is indeed through calcium influx and that calpain activation leads to DLP1 and spectrin cleavage products. Next by using primary neurons from rats as well as CRND8 APP transgenic mice, we showed that this calpain activation and subsequent cleavage of DLP1 is related to Aβ toxicity. Finally, we showed the presence of these cleavage products in AD postmortem patient brains suggesting that aberrant calcium activated calpain is indeed a mechanism in which mitochondrial proteins are cleaved. In summary, we provide a clear mechanism in which

105

Aβ may lead to mitochondrial dysfunction through calpain cleavage of DLP1 and Mfn2 which is observed in both an APP transgenic mouse model as well as in AD patient brains.

Some caveats to this study are revealed in the AD postmortem brain samples where we expected to see a clear increase of the DLP1 calpain cleavage product at 50kDa and even perhaps some at 60kDa, however in addition to the expected significant decrease in full length DLP1, we also observed a significant decrease in the cleavage products. The same phenomenon is present in the spectrin cleavage. This suggests that there are other downstream effectors of the calcium signaling cascade. Recent studies have shown that Aβ-induced calcium influx leads to caspase activation as well as calcineurin activation (D’Amelio et al. 2011; Kamenetz et al. 2003; Shankar et al. 2008;

Wu et al. 2010). It could be possible that further degradation of DLP1 and its cleavage products occurs after the initial cleavage by calpain. Further research needs to be done by looking at the activities of caspase to see if they could also further cleave DLP1 or Mfn2.

It would also be worth finding the exact cleavage sites of DLP1 targeted by calpain in order to try and prevent its cleavage. We first need to establish whether or not DLP1 cleavage by calpain is pathogenic or a neuroprotective consequence since several studies have recently come out suggesting that a reduction in DLP1 levels or activity leads to neuroprotection and reversal of mitochondrial dysfunction (Wang et al. 2017; Kandimalla et al. 2016; Manczak et al. 2016). It is important to look at DLP1 because our previous data has shown that DLP1 alone can rescue the abnormal mitochondrial distribution seen in AD patient fibroblasts (Wang et al. 2008) suggesting that targeting DLP1 levels could provide therapeutic benefit in AD.

106

Conclusion:

In this thesis, we have shown that mitochondrial fragmentation and abnormal distribution causes neurodegeneration in the hippocampus and cortex and with a pathogenesis similar to those pathogenic markers seen in AD such as loss of bioenergetics, oxidative stress, inflammation, and loss of microtubule stability. we have also established a link between Aβ-induced calcium dyshomeostasis and loss of important mitochondrial dynamic proteins and provide extensive support for calpain activation as one mechanism. These data suggest that Aβ-induced mitochondrial disturbances include a myriad of factors and contributions from multiple signaling pathways. Although these pathways must be further explored and elucidated, they provide solid evidence suggesting that indeed abnormal mitochondrial dynamics caused by Aβ or related insult contributes to the mitochondrial dysfunction and ultimately neuronal degeneration in AD brain. This establishes several avenues in which we can pursue therapeutic interventions to halt the development and progression of AD.

107

References

Abramov, A.Y., 2004. Beta-Amyloid Peptides Induce Mitochondrial Dysfunction and Oxidative Stress in Astrocytes and Death of Neurons through Activation of NADPH Oxidase. Journal of Neuroscience, 24(2), pp.565–75. Akulinin, V.A. & Dahlstrom, A., 2003. Quantitative analysis of MAP2 immunoreactivity in human neocortex of three patients surviving after brain ischemia. Neurochemical Research, 28(2), pp.373–378. Alexander, C. et al., 2000. OPA1, encoding a dynamin-related GTPase, is mutated in autosomal dominant optic atrophy linked to chromosome 3q28. Nature Genetics, 26(2), pp.211–5. Alloul, K. et al., 1998. Alzheimer’s disease: A review of the disease, its epidemiology and economic impact. Archives of Gerontology and Geriatrics, pp.189–221. Alonso, A.C. et al., 1994. Role of abnormally phosphorylated tau in the breakdown of microtubules in Alzheimer disease. Proceedings of the National Academy of Sciences of the United States of America, 91(12), pp.5562–5566. Anandatheerthavarada, H.K. et al., 2003. Mitochondrial targeting and a novel transmembrane arrest of Alzheimer’s amyloid precursor protein impairs mitochondrial function in neuronal cells. Journal of Cell Biology, 161, pp.41–54. Atherton, J. et al., 2014. Calpain cleavage and inactivation of the sodium calcium exchanger-3 occur downstream of Aβ in Alzheimer’s disease. Aging Cell, 13(1), pp.49–59. Baek, S.H. et al., 2017. Inhibition of Drp1 Ameliorates Synaptic Depression, Aβ Deposition, and Cognitive Impairment in an Alzheimer’s Disease Model. The Journal of Neuroscience, 37(20), pp.5099–5110. Bagyinszky, E. et al., 2014. The genetics of Alzheimer’s disease. Clinical Interventions in Aging, pp.535–51. Bales, K.R. et al., 2009. Human APOE Isoform-Dependent Effects on Brain beta- amyloid Levels in PDAPP Transgenic Mice. Journal of Neuroscience, 29(21), pp.6771–9. Ban, T. et al., 2017. Molecular basis of selective mitochondrial fusion by heterotypic action between OPA1 and cardiolipin. Nature Cell Biology, 19, p.856. Barsoum, M.J. et al., 2006. Nitric oxide-induced mitochondrial fission is regulated by dynamin-related GTPases in neurons. The EMBO Journal, 25(16), pp.3900–3911. Benard, G. et al., 2007. Mitochondrial bioenergetics and structural network organization. Journal of Cell Science, 120(5), pp.838–48. Berthet, A. et al., 2014. Loss of Mitochondrial Fission Depletes Axonal Mitochondria in Midbrain Dopamine Neurons. Journal of Neuroscience, 34(43), pp.14304–14317. Blass, J.P., 2000. The mitochondrial spiral. An adequate cause of dementia in the Alzheimer’s syndrome. Annals of the New York Academy of Sciences, 924, pp.170– 183.

108

Bleazard, W. et al., 1999. The dynamin-related GTPase Dnm1 regulates mitochondrial fission in yeast. Nature Cell Biology, 1(5), pp.298–304. van der Bliek, A.M., Shen, Q. & Kawajiri, S., 2013. Mechanisms of Mitochondrial Fission and Fusion. Cold Spring Harbor Perspectives in Biology, 5(6), pp.a011072– a011072. Bossy, B. et al., 2010. S-Nitrosylation of DRP1 Does Not Affect Enzymatic Activity and is Not Specific to Alzheimer’s Disease X. Zhu et al., eds. Journal of Alzheimer’s Disease, 20(s2), pp.S513–S526. Braak, H. & Braak, E., 1991. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathologica, 82(4), pp.239–259. Brion, J.-P., 1998. Neurofibrillary Tangles and Alzheimer’s Disease. European Neurology, 40(3), pp.130–140. de Brito, O.M. & Scorrano, L., 2008. Mitofusin 2 tethers endoplasmic reticulum to mitochondria. Nature, 456(7222), pp.605–610. Available at: http://www.nature.com/articles/nature07534. Brookmeyer, R. et al., 2007. Forecasting the global burden of Alzheimer’s disease. Alzheimer’s and Dementia, 3(3), pp.189–91. Calkins, M.J. et al., 2011. Impaired mitochondrial biogenesis, defective axonal transport of mitochondria, abnormal mitochondrial dynamics and synaptic degeneration in a mouse model of Alzheimer’s disease. Human Molecular Genetics, 20(23), pp.4515– 4529. Calkins, M.J. & Reddy, P.H., 2011. Amyloid beta impairs mitochondrial anterograde transport and degenerates synapses in Alzheimer’s disease neurons. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1812(4), pp.507–513. Cardoso, S.M. et al., 2001. Functional mitochondria are required for amyloid beta- mediated neurotoxicity. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 15(8), pp.1439–41. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11387250. Casasnovas, C. et al., 2010. Phenotypic spectrum of MFN2 mutations in the Spanish population. Journal of Medical Genetics, 47(4), pp.249–256. Casley, C.S. et al., 2001. β-Amyloid inhibits integrated mitochondrial respiration and key enzyme activities. Journal of Neurochemistry, 80(1), pp.91–100. Caspersen, C. et al., 2005. Mitochondrial Aβ: a potential focal point for neuronal metabolic dysfunction in Alzheimer’s disease. The FASEB Journal, 19(14), pp.2040–2041. Castellani, R. et al., 2002. Role of mitochondrial dysfunction in Alzheimer’s disease. Journal of Neuroscience Research, 70(3), pp.357–360. Castellano, J.M. et al., 2011. Human apoE isoforms differentially regulate brain amyloid- β peptide clearance. Science Translational Medicine, 3(89), p.89ra57. Cedazo-Minguez, A. et al., 2002. The presenilin 1 deltaE9 mutation gives enhanced basal phospholipase C activity and a resultant increase in intracellular calcium

109

concentrations. The Journal of biological chemistry, 277(39), pp.36646–55. Chan, D.C., 2006. Mitochondria: Dynamic Organelles in Disease, Aging, and Development. Cell, 125(7), pp.1241–1252. Chan, D.C., 2006. Mitochondrial Fusion and Fission in Mammals. Annual Review of Cell and Developmental Biology, 22(1), pp.79–99. Chan, S.L. et al., 2000. Presenilin-1 Mutations Increase Levels of Ryanodine Receptors and Calcium Release in PC12 Cells and Cortical Neurons. Journal of Biological Chemistry, 275(24), pp.18195–18200. Chandhok, G., Lazarou, M. & Neumann, B., 2018. Structure, function, and regulation of mitofusin-2 in health and disease. Biological Reviews, 93(2), pp.933–949. Chang, C.-R. & Blackstone, C., 2007. Cyclic AMP-dependent Protein Kinase Phosphorylation of Drp1 Regulates Its GTPase Activity and Mitochondrial Morphology. Journal of Biological Chemistry, 282(30), pp.21583–21587. Chen, H. et al., 2010. Mitochondrial Fusion Is Required for mtDNA Stability in Skeletal Muscle and Tolerance of mtDNA Mutations. Cell, 141(2), pp.280–289. Chen, H. et al., 2003. Mitofusins Mfn1 and Mfn2 coordinately regulate mitochondrial fusion and are essential for embryonic development. Journal of Cell Biology, 160(2), pp.189–200. Chen, H., Chomyn, A. & Chan, D.C., 2005. Disruption of fusion results in mitochondrial heterogeneity and dysfunction. Journal of Biological Chemistry, 280(28), pp.26185– 26192. Chen, H., McCaffery, J.M. & Chan, D.C., 2007. Mitochondrial Fusion Protects against Neurodegeneration in the Cerebellum. Cell, 130(3), pp.548–562. Chishti, M.A. et al., 2001. Early-onset Amyloid Deposition and Cognitive Deficits in Transgenic Mice Expressing a Double Mutant Form of Amyloid Precursor Protein 695. Journal of Biological Chemistry, 276(24), pp.21562–21570. Cho, D.-H. et al., 2009. S-nitrosylation of Drp1 mediates beta-amyloid-related mitochondrial fission and neuronal injury. Science (New York, N.Y.), 324, pp.102– 105. Chui, D.-H. et al., 1999. Transgenic mice with Alzheimer presenilin 1 mutations show accelerated neurodegeneration without amyloid plaque formation. Nature Medicine, 5(5), pp.560–564. Cogliati, S. et al., 2013. Mitochondrial Cristae Shape Determines Respiratory Chain Supercomplexes Assembly and Respiratory Efficiency. Cell, 155(1), pp.160–171. Cook, D.G. et al., 1997. Alzheimer’s A beta(1-42) is generated in the endoplasmic reticulum/intermediate compartment of NT2N cells. Nature medicine, 3(9), pp.1021–3. Corder, E.H. et al., 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science, 261(5123), pp.921–3. Corral-Debrinski, M. et al., 1994. Marked changes in mitochondrial dna deletion levels in alzheimer brains. Genomics, 23(2), pp.471–6. 110

Coskun, P.E., Beal, M.F. & Wallace, D.C., 2004. Alzheimer’s brains harbor somatic mtDNA control-region mutations that suppress mitochondrial transcription and replication. Proceedings of the National Academy of Sciences, 101(29), pp.10726– 31. Cottrell, D.A. et al., 2001. Mitochondrial enzyme-deficient hippocampal neurons and choroidal cells in AD. Neurology, 57(2), pp.260–4. Cribbs, J.T. & Strack, S., 2007. Reversible phosphorylation of Drp1 by cyclic AMP- dependent protein kinase and calcineurin regulates mitochondrial fission and cell death. EMBO reports, 8(10), pp.939–944. Cunnane, S.C. & Crawford, M.A., 2014. Energetic and nutritional constraints on infant brain development: Implications for brain expansion during human evolution. Journal of Human Evolution, 77, pp.88–98. D’Amelio, M. et al., 2011. Caspase-3 triggers early synaptic dysfunction in a mouse model of Alzheimer’s disease. Nature Neuroscience, 14(1), pp.69–76. Dahlgren, K.N. et al., 2002. Oligomeric and fibrillar species of amyloid-β peptides differentially affect neuronal viability. Journal of Biological Chemistry, 277(35), pp.32046–53. Davolio, C. & Greenamyre, J.T., 1995. Selective vulnerability of the CA1 region of hippocampus to the indirect excitotoxic effects of malonic acid. Neuroscience Letters, 192(1), pp.29–32. Decker, H. et al., 2010. Amyloid-beta Peptide Oligomers Disrupt Axonal Transport through an NMDA Receptor-Dependent Mechanism That Is Mediated by Glycogen Synthase Kinase 3 in Primary Cultured Hippocampal Neurons. Journal of Neuroscience, 30(27), pp.9166–71. Delettre, C. et al., 2000. Nuclear gene OPA1, encoding a mitochondrial dynamin-related protein, is mutated in dominant optic atrophy. Nature Genetics, 26(2), pp.207–10. Dougherty, J.J., Wu, J. & Nichols, R. a, 2003. Beta-amyloid regulation of presynaptic nicotinic receptors in rat hippocampus and neocortex. The Journal of neuroscience : the official journal of the Society for Neuroscience, 23(17), pp.6740–7. Du, H. et al., 2011. Cyclophilin D deficiency improves mitochondrial function and learning/memory in aging Alzheimer disease mouse model. Neurobiology of Aging, 32(3), pp.398–406. Du, H. et al., 2010. Early deficits in synaptic mitochondria in an Alzheimer’s disease mouse model. Proceedings of the National Academy of Sciences, 107(43), pp.18670–18675. DuBoff, B., Götz, J. & Feany, M.B., 2012. Tau Promotes Neurodegeneration via DRP1 Mislocalization In Vivo. Neuron, 75(4), pp.618–632. Elgass, K. et al., 2013. Recent advances into the understanding of mitochondrial fission. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1833(1), pp.150– 161. Estaquier, J. & Arnoult, D., 2007. Inhibiting Drp1-mediated mitochondrial fission

111

selectively prevents the release of cytochrome c during apoptosis. Cell Death & Differentiation, 14(6), pp.1086–1094. Faelber, K. et al., 2011. Crystal structure of nucleotide-free dynamin. Nature, 477(7366), pp.556–560. Federico, A. et al., 2012. Mitochondria, oxidative stress and neurodegeneration. Journal of the Neurological Sciences, 322(1–2), pp.254–262. Ferreira, A., 2012. Calpain Dysregulation in Alzheimer’s Disease. ISRN Biochemistry, 2012, pp.1–12. Ferreira, A. & Bigio, E., 2011. Calpain-mediated tau cleavage: a mechanism leading to neurodegeneration shared by multiple tauopathies. Molecular Medicine, 17(7–8), p.1. Ferrer, I. et al., 2005. Current Advances on Different Kinases Involved in Tau Phosphorylation, and Implications in Alzheimers Disease and Tauopathies. Current Alzheimer Research, 2(1), pp.3–18. Fifre, A. et al., 2006. Microtubule-associated protein MAP1A, MAP1B, and MAP2 proteolysis during soluble amyloid ??-peptide-induced neuronal apoptosis: Synergistic involvement of calpain and caspase-3. Journal of Biological Chemistry, 281(1), pp.229–240. Ford, M.G.J., Jenni, S. & Nunnari, J., 2011. The crystal structure of dynamin. Nature, 477(7366), pp.561–566. Förstl, H. & Kurz, A., 1999. Clinical features of Alzheimer’s disease. European Archives of Psychiatry and Clinical Neuroscience, 249(6), pp.288–290. Fox, T.D., 2012. An MBoC Favorite: Mitochondrial transmission during mating in Saccharomyces cerevisiae is determined by mitochondrial fusion and fission and the intramitochondrial segregation of mitochondrial DNA. Molecular Biology of the Cell, 23(21), pp.4144–4144. Franco, A. et al., 2016. Correcting mitochondrial fusion by manipulating mitofusin conformations. Nature, 540(7631), pp.74–79. Frank, S. et al., 2001. The role of dynamin-related protein 1, a mediator of mitochondrial fission, in apoptosis. Developmental cell, 1(4), pp.515–25. Freir, D.B. et al., 2011. Interaction between prion protein and toxic amyloid β assemblies can be therapeutically targeted at multiple sites. Nature Communications, 2(1), p.336. Frey, T.G. & Mannella, C.A., 2000. The internal structure of mitochondria. Trends in biochemical sciences, 25(7), pp.319–24. Fröhlich, C. et al., 2013. Structural insights into oligomerization and mitochondrial remodelling of dynamin 1-like protein. EMBO Journal, 32(9), pp.1280–92. Gan, X. et al., 2014. Inhibition of ERK-DLP1 signaling and mitochondrial division alleviates mitochondrial dysfunction in Alzheimer’s disease cybrid cell. Biochimica et Biophysica Acta - Molecular Basis of Disease, 1842(2), pp.220–231. Garg, S. et al., 2011. Cleavage of Tau by calpain in Alzheimer’s disease: The quest for 112

the toxic 17 kD fragment. Neurobiology of Aging, 32(1), pp.1–14. Gibson, G.E., Sheu, K.-F.R. & Blass, J.P., 1998. Abnormalities of mitochondrial enzymes in Alzheimer disease. Journal of Neural Transmission, 105(8–9), pp.855– 870. Goedert, M. et al., 1989. Multiple isoforms of human microtubule-associated protein tau: sequences and localization in neurofibrillary tangles of Alzheimer’s disease. Neuron, 3(4), pp.519–26. Granic, I. et al., 2010. Calpain inhibition prevents amyloid-β-induced neurodegeneration and associated behavioral dysfunction in rats. Neuropharmacology, 59, pp.334–342. Harder, Z., Zunino, R. & McBride, H., 2004. Sumo1 Conjugates Mitochondrial Substrates and Participates in Mitochondrial Fission. Current Biology, 14(4), pp.340–345. Hardy, J. & Higgins, G., 1992. Alzheimer’s disease: the amyloid cascade hypothesis. Science, 256(5054), pp.184–185. Hartmann, T. et al., 1997. Distinct sites of intracellular production for Alzheimer’s disease A beta40/42 amyloid peptides. Nature medicine, 3(9), pp.1016–20. Hirai, K. et al., 2001. Mitochondrial abnormalities in Alzheimer’s disease. The Journal of neuroscience : the official journal of the Society for Neuroscience, 21(9), pp.3017– 23. Hoekstra, J.G. et al., 2016. Mitochondrial DNA mutations increase in early stage Alzheimer disease and are inconsistent with oxidative damage. Annals of neurology, 80(2), pp.301–6. Iijima-Ando, K. et al., 2009. Mitochondrial Mislocalization Underlies Aβ42-Induced Neuronal Dysfunction in a Drosophila Model of Alzheimer’s Disease M. B. Feany, ed. PLoS ONE, 4(12), p.e8310. Iqbal, K. et al., 2010. Tau in Alzheimer disease and related tauopathies. Current Alzheimer research, 7(8), pp.656–64. Ishihara, N., Eura, Y. & Mihara, K., 2004. Mitofusin 1 and 2 play distinct roles in mitochondrial fusion reactions via GTPase activity. Journal of cell science, 117(Pt 26), pp.6535–46. James, D.I. et al., 2003. hFis1, a novel component of the mammalian mitochondrial fission machinery. The Journal of biological chemistry, 278(38), pp.36373–9. Jiang, S. et al., 2018. Mfn2 ablation causes an oxidative stress response and eventual neuronal death in the hippocampus and cortex. Molecular Neurodegeneration, 13(1), p.5. Jin, N. et al., 2015. Truncation and activation of GSK-3β by calpain I: a molecular mechanism links to tau hyperphosphorylation in Alzheimer’s disease. Scientific Reports, 5(1), p.8187. Jones, L. et al., 2015. Convergent genetic and expression data implicate immunity in Alzheimer’s disease. Alzheimer’s & Dementia, 11(6), pp.658–671. De Jonghe, C., 2001. Pathogenic APP mutations near the gamma-secretase cleavage site 113

differentially affect Abeta secretion and APP C-terminal fragment stability. Human Molecular Genetics, 10(16), pp.1665–1671. Kaech, S. & Banker, G., 2006. Culturing hippocampal neurons. Nature Protocols, 1(5), pp.2406–2415. Kamat, P.K., Rai, S. & Nath, C., 2013. Okadaic acid induced neurotoxicity: An emerging tool to study Alzheimer’s disease pathology. NeuroToxicology, 37, pp.163–172. Kamenetz, F. et al., 2003. APP processing and synaptic function. Neuron, 37(6), pp.925– 37. Kandimalla, R. et al., 2016. Reduced dynamin-related protein 1 protects against phosphorylated Tau-induced mitochondrial dysfunction and synaptic damage in Alzheimer’s disease. Human Molecular Genetics, p.ddw312. Kann, O. & Kovács, R., 2007. Mitochondria and neuronal activity. American journal of physiology. Cell physiology, 292, pp.C641–C657. Kayed, R. & Lasagna-Reeves, C.A., 2012. Molecular Mechanisms of Amyloid Oligomers Toxicity G. Perry et al., eds. Journal of Alzheimer’s Disease, 33(s1), pp.S67–S78. Kelly, B.L. & Ferreira, A., 2006. β-Amyloid-induced Dynamin 1 Degradation Is Mediated by N -Methyl-D-Aspartate Receptors in Hippocampal Neurons. Journal of Biological Chemistry, 281(38), pp.28079–28089. Kelly, B.L., Vassar, R. & Ferreira, A., 2005. β-amyloid-induced dynamin 1 depletion in hippocampal neurons: A potential mechanism for early cognitive decline in Alzheimer disease. Journal of Biological Chemistry, 280, pp.31746–31753. Khan, S.M. et al., 2000. Alzheimer’s disease cybrids replicate beta-amyloid abnormalities through cell death pathways. Annals of neurology, 48(2), pp.148–55. Kijima, K. et al., 2005. Mitochondrial GTPase mitofusin 2 mutation in Charcot-Marie- Tooth neuropathy type 2A. Human genetics, 116(1–2), pp.23–7. Kim, J., Basak, J.M. & Holtzman, D.M., 2009. The Role of Apolipoprotein E in Alzheimer’s Disease. Neuron, 63(3), pp.287–303. Kirkitadze, M.D. & Kowalska, A., 2005. Molecular mechanisms initiating amyloid beta- fibril formation in Alzheimer’s disease. Acta biochimica Polonica, 52(2), pp.417– 23. Knobloch, M. et al., 2007. Intracellular Aβ and cognitive deficits precede β-amyloid deposition in transgenic arcAβ mice. Neurobiology of Aging, 28(9), pp.1297–1306. Knott, A.B. et al., 2008. Mitochondrial fragmentation in neurodegeneration. Nature Reviews Neuroscience, 9(7), pp.505–518. König, G. et al., 1992. Identification of a novel alternative splice isoform of the βA4 amyloid precursor protein (APP) MRNA in leukocytes and brain microglial cells. Neurobiology of Aging, 13, p.S68. Koshiba, T., 2004. Structural Basis of Mitochondrial Tethering by Mitofusin Complexes. Science, 305(5685), pp.858–862.

114

Kumar, S. et al., 2016. Activation of Mitofusin2 by Smad2-RIN1 Complex during Mitochondrial Fusion. Molecular Cell, 62(4), pp.520–531. Kwei, S., Jiang, C. & Haddad, G.G., 1993. Acute anoxia-induced alterations in MAP2 immunoreactivity and neuronal morphology in rat hippocampus. Brain Research, 620(2), pp.203–210. LaFerla, F.M., 2002. Calcium dyshomeostasis and intracellular signalling in alzheimer’s disease. Nat Rev Neurosci, 3(11), pp.862–872. LaFerla, F.M., Green, K.N. & Oddo, S., 2007. Intracellular amyloid-β in Alzheimer’s disease. Nature Reviews Neuroscience, 8(7), pp.499–509. Larson, E.B. et al., 2004. Survival after Initial Diagnosis of Alzheimer Disease. Annals of Internal Medicine, 140(7), p.501. Laurén, J. et al., 2009. Cellular prion protein mediates impairment of synaptic plasticity by amyloid-β oligomers. Nature, 457(7233), pp.1128–1132. Leal, N.S. et al., 2016. Mitofusin‐2 knockdown increases ER–mitochondria contact and decreases amyloid β‐peptide production. Journal of Cellular and Molecular Medicine, 20(9), pp.1686–1695. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4988279/. Lee, S. et al., 2012. Mitofusin 2 is necessary for striatal axonal projections of midbrain dopamine neurons. Human Molecular Genetics, 21(22), pp.4827–4835. Li, Q.-X. et al., 2002. Intracellular Accumulation of Detergent-Soluble Amyloidogenic Aβ Fragment of Alzheimer’s Disease Precursor Protein in the Hippocampus of Aged Transgenic Mice. Journal of Neurochemistry, 72(6), pp.2479–2487. Li, S. et al., 2011. Soluble Aβ oligomers inhibit long-term potentiation through a mechanism involving excessive activation of extrasynaptic NR2B-containing NMDA receptors. The Journal of neuroscience : the official journal of the Society for Neuroscience, 31(18), pp.6627–38. Li, X.-C. et al., 2016. Human wild-type full-length tau accumulation disrupts mitochondrial dynamics and the functions via increasing mitofusins. Scientific Reports, 6(1), p.24756. Liang, B. et al., 2010. Calpain Activation Promotes BACE1 Expression, Amyloid Precursor Protein Processing, and Amyloid Plaque Formation in a Transgenic Mouse Model of Alzheimer Disease. Journal of Biological Chemistry, 285(36), pp.27737–27744. Liang, Z. et al., 2008. Decrease of Protein Phosphatase 2A and its Association with Accumulation and Hyperphosphorylation of Tau in Down Syndrome. Journal of Alzheimer’s Disease, 13(3), pp.295–302. Lindwall, G. & Cole, R.D., 1984. Phosphorylation affects the ability of tau protein to promote microtubule assembly. The Journal of biological chemistry, 259(8), pp.5301–5. Liu, F. et al., 2005. Truncation and Activation of Calcineurin A by Calpain I in Alzheimer Disease Brain. Journal of Biological Chemistry, 280(45), pp.37755–

115

37762. Liu, Q. et al., 2005. Alzheimer-specific epitopes of tau represent lipid peroxidation- induced conformations. Free Radical Biology and Medicine, 38(6), pp.746–754. Liu, W. et al., 2011. Pink1 regulates the oxidative phosphorylation machinery via mitochondrial fission. Proceedings of the National Academy of Sciences, 108(31), pp.12920–12924. Loiseau, D. et al., 2007. Mitochondrial coupling defect in Charcot-Marie-Tooth type 2A disease. Annals of Neurology, 61(4), pp.315–323. Losón, O.C. et al., 2013. Fis1, Mff, MiD49, and MiD51 mediate Drp1 recruitment in mitochondrial fission D. D. Newmeyer, ed. Molecular Biology of the Cell, 24(5), pp.659–667. Loucks, F.A. et al., 2009. Caspases indirectly regulate cleavage of the mitochondrial fusion GTPase OPA1 in neurons undergoing apoptosis. Brain Research, 1250, pp.63–74. MacAskill, A.F. et al., 2009. Miro1 Is a Calcium Sensor for Glutamate Receptor- Dependent Localization of Mitochondria at Synapses. Neuron, 61(4), pp.541–555. Mahley, R.W., Weisgraber, K.H. & Huang, Y., 2009. Apolipoprotein E: structure determines function, from atherosclerosis to Alzheimer’s disease to AIDS. Journal of Lipid Research, 50(Supplement), pp.S183–S188. Mahley, R.W., Weisgraber, K.H. & Huang, Y., 2006. Apolipoprotein E4: A causative factor and therapeutic target in neuropathology, including Alzheimer’s disease. Proceedings of the National Academy of Sciences, 103(15), pp.5644–5651. Manczak, M. et al., 2010. Mitochondria-Targeted Antioxidants Protect Against Amyloid- β Toxicity in Alzheimer’s Disease Neurons X. Zhu et al., eds. Journal of Alzheimer’s Disease, 20(s2), pp.S609–S631. Manczak, M. et al., 2006. Mitochondria are a direct site of Aβ accumulation in Alzheimer’s disease neurons: implications for free radical generation and oxidative damage in disease progression. Human Molecular Genetics, 15(9), pp.1437–1449. Manczak, M. et al., 2016. Protective effects of reduced dynamin-related protein 1 against amyloid beta-induced mitochondrial dysfunction and synaptic damage in Alzheimer’s disease. Human Molecular Genetics, p.ddw330. Manczak, M., Calkins, M.J. & Reddy, P.H., 2011. Impaired mitochondrial dynamics and abnormal interaction of amyloid beta with mitochondrial protein Drp1 in neurons from patients with Alzheimer’s disease: implications for neuronal damage. Human Molecular Genetics, 20(13), pp.2495–2509. Manczak, M. & Reddy, P.H., 2012. Abnormal interaction between the mitochondrial fission protein Drp1 and hyperphosphorylated tau in Alzheimer’s disease neurons: implications for mitochondrial dysfunction and neuronal damage. Human Molecular Genetics, 21(11), pp.2538–2547. Mattson, M.P. et al., 2000. Calcium signaling in the ER: its role in neuronal plasticity and neurodegenerative disorders. Trends in Neurosciences, 23(5), pp.222–229.

116

Mattson, M.P., 2004. Pathways towards and away from Alzheimer’s disease. Nature, 430(7000), pp.631–639. Mattson, M.P. & Furukawa, K., 1997. Anti-apoptotic actions of cycloheximide: blockade of programmed cell death or induction of programmed cell life? Apoptosis, 2(3), pp.257–264. Mattson, M.P., Gleichmann, M. & Cheng, A., 2008. Mitochondria in Neuroplasticity and Neurological Disorders. Neuron, 60(5), pp.748–766. Maurer, I., 2000. A selective defect of cytochrome c oxidase is present in brain of Alzheimer disease patients. Neurobiology of Aging, 21(3), pp.455–462. Mazanetz, M.P. & Fischer, P.M., 2007. Untangling tau hyperphosphorylation in drug design for neurodegenerative diseases. Nature Reviews Drug Discovery, 6(6), pp.464–479. McKhann, G. et al., 1984. Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group* under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34(7), pp.939–939. McLelland, G.-L. et al., 2018. Mfn2 ubiquitination by PINK1/parkin gates the p97- dependent release of ER from mitochondria to drive mitophagy I. Dikic, ed. eLife, 7, p.e32866. Mears, J.A. et al., 2011. Conformational changes in Dnm1 support a contractile mechanism for mitochondrial fission. Nature Structural & Molecular Biology, 18(1), pp.20–26. Migliore, L. et al., 2005. Oxidative DNA damage in peripheral leukocytes of mild cognitive impairment and AD patients. Neurobiology of Aging, 26(5), pp.567–573. Minoshima, S. et al., 1997. Metabolic reduction in the posterior cingulate cortex in very early Alzheimer’s disease. Annals of Neurology, 42(1), pp.85–94. Available at: http://doi.wiley.com/10.1002/ana.410420114. Mishra, P. & Chan, D.C., 2014. Mitochondrial dynamics and inheritance during cell division, development and disease. Nature Reviews Molecular Cell Biology, 15(10), pp.634–46. Misko, a. L. et al., 2012. Mitofusin2 Mutations Disrupt Axonal Mitochondrial Positioning and Promote Axon Degeneration. Journal of Neuroscience, 32(12), pp.4145–4155. Misko, A. et al., 2010. Mitofusin 2 is necessary for transport of axonal mitochondria and interacts with the Miro/Milton complex. The Journal of neuroscience : the official journal of the Society for Neuroscience, 30(12), pp.4232–4240. Morales-Corraliza, J. et al., 2012. Calpastatin modulates APP processing in the brains of β-amyloid depositing but not wild-type mice. Neurobiology of Aging, 33(6), p.1125.e9-1125.e18. Morkuniene, R. et al., 2015. Small Aβ 1-42 oligomer-induced membrane depolarization of neuronal and microglial cells: Role of N-methyl-D-aspartate receptors. Journal of Neuroscience Research, 93(3), pp.475–486.

117

Nagy, Z. et al., 1999. Mitochondrial enzyme expression in the hippocampus in relation to Alzheimer-type pathology. Acta Neuropathologica, 97(4), pp.346–354. Nemani, N. et al., 2018. MIRO-1 Determines Mitochondrial Shape Transition upon GPCR Activation and Ca 2+ Stress. Cell Reports, 23(4), pp.1005–1019. Newman, M., Musgrave, F.I. & Lardelli, M., 2007. Alzheimer disease: Amyloidogenesis, the presenilins and animal models. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1772(3), pp.285–297. Nochez, Y. et al., 2009. Acute and late-onset optic atrophy due to a novel OPA1 mutation leading to a mitochondrial coupling defect. Molecular vision, 15, pp.598–608. Nunomura, A. et al., 2006. Involvement of Oxidative Stress in Alzheimer Disease. Journal of Neuropathology and Experimental Neurology, 65(7), pp.631–641. Nunomura, A. et al., 2012. The Earliest Stage of Cognitive Impairment in Transition From Normal Aging to Alzheimer Disease Is Marked by Prominent RNA Oxidation in Vulnerable Neurons. Journal of Neuropathology & Experimental Neurology, 71(3), pp.233–241. Olichon, A. et al., 2003. Loss of OPA1 Perturbates the Mitochondrial Inner Membrane Structure and Integrity, Leading to Cytochrome c Release and Apoptosis. Journal of Biological Chemistry, 278(10), pp.7743–7746. Otera, H. et al., 2016. Drp1-dependent mitochondrial fission via MiD49/51 is essential for apoptotic cristae remodeling. The Journal of Cell Biology, 212(5), pp.531–544. Otera, H., Ishihara, N. & Mihara, K., 2013. New insights into the function and regulation of mitochondrial fission. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, pp.1256–1268. Pahl, H.L. & Baeuerle, P.A., 1996. Activation of NF-κB by ER stress requires both Ca 2+ and reactive oxygen intermediates as messengers. FEBS Letters, 392(2), pp.129– 136. Pallanck, L., 2013. Mitophagy: Mitofusin Recruits a Mitochondrial Killer. Current Biology, 23(13), pp.R570–R572. Paltauf-Doburzynska, J., Malli, R. & Graier, W.F., 2004. Hyperglycemic Conditions Affect Shape and Ca2+ Homeostasis of Mitochondria in Endothelial Cells. Journal of Cardiovascular Pharmacology, 44(4), pp.423–436. Pang, Z., Umberger, G.H. & Geddes, J.W., 1996. Neuronal loss and cytoskeletal disruption following intrahippocampal administration of the metabolic inhibitor malonate: lack of protection by MK-801. J Neurochem, 66(2), pp.474–484. Paolicelli, R.C. et al., 2017. TDP-43 Depletion in Microglia Promotes Amyloid Clearance but Also Induces Synapse Loss. Neuron, 95(2), p.297–308.e6. Park, J. et al., 2015. Loss of mitofusin 2 links beta-amyloid-mediated mitochondrial fragmentation and Cdk5-induced oxidative stress in neuron cells. Journal of Neurochemistry, 132(6), pp.687–702. Pérez, M. et al., 2000. Phosphorylated, but not native, tau protein assembles following reaction with the lipid peroxidation product, 4-hydroxy-2-nonenal. FEBS Letters,

118

486(3), pp.270–274. Available at: http://dx.doi.org/10.1016/S0014-5793(00)02323- 1. Pericak-Vance, M.A. et al., 1991. Linkage studies in familial Alzheimer disease: evidence for chromosome 19 linkage. American journal of human genetics, 48(6), pp.1034–50. Pham, A.H. et al., 2012. Loss of Mfn2 results in progressive, retrograde degeneration of dopaminergic neurons in the nigrostriatal circuit. Human Molecular Genetics, 21(22), pp.4817–4826. Picard, M. et al., 2013. Mitochondrial morphology transitions and functions: implications for retrograde signaling? American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 304(6), pp.R393–R406. Plassman, B.L. et al., 2007. Prevalence of Dementia in the United States: The Aging, Demographics, and Memory Study. Neuroepidemiology, 29(1–2), pp.125–132. Prasher, V.P. et al., 1998. Molecular mapping of alzheimer-type dementia in Down’s syndrome. Annals of Neurology, 43(3), pp.380–383. Rajgopal, Y. & Vemuri, M.C., 2002. Calpain activation and α-spectrin cleavage in rat brain by ethanol. Neuroscience Letters, 321(3), pp.187–91. Reck, G. & Mach, H., 1988. Untersuchungen zur Regulation der fetoplazentaren Einheit. Geburtshilfe und Frauenheilkunde, 48(09), pp.651–653. Reddy, P.H. et al., 2012. Abnormal mitochondrial dynamics and synaptic degeneration as early events in Alzheimer’s disease: Implications to mitochondria-targeted antioxidant therapeutics. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1822(5), pp.639–649. Reddy, P.H. et al., 2004. Gene expression profiles of transcripts in amyloid precursor protein transgenic mice: up-regulation of mitochondrial metabolism and apoptotic genes is an early cellular change in Alzheimer’s disease. Human Molecular Genetics, 13(12), pp.1225–1240. Reiman, E.M. et al., 2009. Fibrillar amyloid- burden in cognitively normal people at 3 levels of genetic risk for Alzheimer’s disease. Proceedings of the National Academy of Sciences, 106(16), pp.6820–6825. Roberson, E.D. et al., 2007. Reducing Endogenous Tau Ameliorates Amyloid -Induced Deficits in an Alzheimer’s Disease Mouse Model. Science, 316(5825), pp.750–754. Rui, Y. et al., 2006. Acute impairment of mitochondrial trafficking by beta-amyloid peptides in hippocampal neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience, 26(41), pp.10480–10487. Saito, K. et al., 1993. Widespread activation of calcium-activated neutral proteinase (calpain) in the brain in Alzheimer disease: a potential molecular basis for neuronal degeneration. Proceedings of the National Academy of Sciences of the United States of America, 90(7), pp.2628–32. Salthouse, T.A., 2004. What and When of Cognitive Aging. Current Directions in Psychological Science, 13(4), pp.140–144.

119

Schmitt, H., Gozes, I. & Littauer, U.Z., 1977. Decrease in levels and rates of synthesis of tubulin and actin in developing rat brain. Brain Research, 121(2), pp.327–342. Schober, J.M. et al., 2012. The microtubule-associated protein EB1 maintains cell polarity through activation of protein kinase C. Biochemical and Biophysical Research Communications, 417(1), pp.67–72. Schönfeld, P. & Reiser, G., 2017. Brain energy metabolism spurns fatty acids as fuel due to their inherent mitotoxicity and potential capacity to unleash neurodegeneration. Neurochemistry International, 109, pp.68–77. Sengupta, U., Nilson, A.N. & Kayed, R., 2016. The Role of Amyloid-β Oligomers in Toxicity, Propagation, and Immunotherapy. EBioMedicine, 6, pp.42–49. Sesaki, H. et al., 2014. In vivo functions of Drp1: Lessons learned from yeast genetics and mouse knockouts. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1842(8), pp.1179–1185. Sesaki, H. & Jensen, R.E., 1999. Division versus Fusion: Dnm1p and Fzo1p Antagonistically Regulate Mitochondrial Shape. The Journal of Cell Biology, 147(4), pp.699–706. Shankar, G.M. et al., 2008. Amyloid-β protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory. Nature Medicine, 14(8), pp.837–842. Shields, L.Y. et al., 2015. Dynamin-related protein 1 is required for normal mitochondrial bioenergetic and synaptic function in CA1 hippocampal neurons. Cell Death & Disease, 6(4), pp.e1725–e1725. Silva, D.F. et al., 2012. Mitochondrial Abnormalities in Alzheimer’s Disease. In Advances in pharmacology (San Diego, Calif.). pp. 83–126. Simpkins, K.L. et al., 2003. Selective activation induced cleavage of the NR2B subunit by calpain. The Journal of neuroscience : the official journal of the Society for Neuroscience, 23(36), pp.11322–31. Smirnova, E. et al., 1998. A Human Dynamin-related Protein Controls the Distribution of Mitochondria. The Journal of Cell Biology, 143(2), pp.351–358. Smirnova, E. et al., 2001. Dynamin-related Protein Drp1 Is Required for Mitochondrial Division in Mammalian Cells T. D. Pollard, ed. Molecular Biology of the Cell, 12(8), pp.2245–2256. Smith, I.F. et al., 2001. Effects of chronic hypoxia on Ca(2+) stores and capacitative Ca(2+) entry in human neuroblastoma (SH-SY5Y) cells. Journal of neurochemistry, 79(4), pp.877–84. Smith, I.F. et al., 2005. Enhanced caffeine-induced Ca2+ release in the 3xTg-AD mouse model of Alzheimer’s disease. Journal of Neurochemistry, 94(6), pp.1711–1718. Smith, M.A., 1998. Alzheimer Disease. In International Review of Neurobiology. Academic Press, pp. 1–54. Song, H.L. et al., 2014. β-Amyloid is transmitted via neuronal connections along axonal membranes. Annals of Neurology, 75(1), pp.88–97. Song, Z. et al., 2007. OPA1 processing controls mitochondrial fusion and is regulated by 120

mRNA splicing, membrane potential, and Yme1L. The Journal of Cell Biology, 178(5), pp.749–755. Spinazzi, M. et al., 2008. A novel deletion in the GTPase domain of OPA1 causes defects in mitochondrial morphology and distribution, but not in function. Human Molecular Genetics, 17(21), pp.3291–3302. Stokin, G.B., 2005. Axonopathy and Transport Deficits Early in the Pathogenesis of Alzheimer’s Disease. Science, 307(5713), pp.1282–1288. Stutzmann, G.E., 2007. The Pathogenesis of Alzheimers Disease—Is It a Lifelong “Calciumopathy”? The Neuroscientist, 13(5), pp.546–559. Su, B. et al., 2010. Chronic oxidative stress causes increased tau phosphorylation in M17 neuroblastoma cells. Neuroscience letters, 468(3), pp.267–271. Su, B. et al., 2008. Physiological regulation of tau phosphorylation during hibernation. Journal of Neurochemistry, 105(6), pp.2098–2108. Sugioka, R., Shimizu, S. & Tsujimoto, Y., 2004. Fzo1, a Protein Involved in Mitochondrial Fusion, Inhibits Apoptosis. Journal of Biological Chemistry, 279(50), pp.52726–52734. Sullivan, P.G. & Brown, M.R., 2005. Mitochondrial aging and dysfunction in Alzheimer’s disease. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 29(3), pp.407–410. Swerdlow, R.H., 2016. Bioenergetics and metabolism: a bench to bedside perspective. Journal of Neurochemistry, 139, pp.126–135. Swerdlow, R.H., 2007. Is aging part of Alzheimer’s disease, or is Alzheimer’s disease part of aging? Neurobiology of Aging, 28(10), pp.1465–1480. Swerdlow, R.H. & Khan, S.M., 2004. A “mitochondrial cascade hypothesis” for sporadic Alzheimer’s disease. Medical Hypotheses, 63(1), pp.8–20. Szabadkai, G. et al., 2006. Mitochondrial dynamics and Ca2+ signaling. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1763(5–6), pp.442–449. Taguchi, N. et al., 2007. Mitotic phosphorylation of dynamin-related GTPase Drp1 participates in mitochondrial fission. Journal of Biological Chemistry, 282(15), pp.11521–11529. Takeda, A. et al., 2000. In Alzheimer’s disease, heme oxygenase is coincident with Alz50, an epitope of τ induced by 4-hydroxy-2-nonenal modification. Journal of Neurochemistry, 75(3), pp.1234–1241. Tang, Y.-P. & Gershon, E.S., 2003. Genetic studies in Alzheimer’s disease. Dialogues in clinical neuroscience, 5(1), pp.17–26. Trushina, E. et al., 2012. Defects in Mitochondrial Dynamics and Metabolomic Signatures of Evolving Energetic Stress in Mouse Models of Familial Alzheimer’s Disease D. I. Finkelstein, ed. PLoS ONE, 7(2), p.e32737. Twig, G. et al., 2008. Fission and selective fusion govern mitochondrial segregation and elimination by autophagy. The EMBO Journal, 27(2), pp.433–446.

121

Ueda, K. et al., 2002. Amyloid β Protein Potentiates Ca2+ Influx Through L-Type Voltage-Sensitive Ca2+ Channels: A Possible Involvement of Free Radicals. Journal of Neurochemistry, 68(1), pp.265–271. Um, J.W. et al., 2012. Alzheimer amyloid-β oligomer bound to postsynaptic prion protein activates Fyn to impair neurons. Nature Neuroscience, 15(9), pp.1227–1235. Wakabayashi, J. et al., 2009. The dynamin-related GTPase Drp1 is required for embryonic and brain development in mice. The Journal of Cell Biology, 186(6), pp.805–816. Wang, W. et al., 2017. Inhibition of mitochondrial fragmentation protects against Alzheimer’s disease in rodent model. Human Molecular Genetics, 26(21), pp.4118– 4131. Wang, W. et al., 2015. MFN2 Couples Glutamate Excitotoxicity and Mitochondrial Dysfunction in Motor Neurons. Journal of Biological Chemistry, 290(1), pp.168– 182. Wang, W. et al., 2016. The inhibition of TDP-43 mitochondrial localization blocks its neuronal toxicity. Nature Medicine, 22(8), pp.869–878. Wang, X. et al., 2008. Amyloid- overproduction causes abnormal mitochondrial dynamics via differential modulation of mitochondrial fission/fusion proteins. Proceedings of the National Academy of Sciences, 105(49), pp.19318–19323. Wang, X. et al., 2010. Amyloid-β-Derived Diffusible Ligands Cause Impaired Axonal Transport of Mitochondria in Neurons. Neurodegenerative Diseases, 7(1–3), pp.56– 59. Wang, X. et al., 2008. Dynamin-Like Protein 1 Reduction Underlies Mitochondrial Morphology and Distribution Abnormalities in Fibroblasts from Sporadic Alzheimer’s Disease Patients. The American Journal of Pathology, 173(2), pp.470– 482. Wang, X. et al., 2009. Impaired Balance of Mitochondrial Fission and Fusion in Alzheimer’s Disease. Journal of Neuroscience, 29(28), pp.9090–9103. Wang, X. et al., 2009. Impaired Balance of Mitochondrial Fission and Fusion in Alzheimer’s Disease. Journal of Neuroscience, 29(28), pp.9090–9103. Wang, X. et al., 2014. Oxidative stress and mitochondrial dysfunction in Alzheimer’s disease. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1842(8), pp.1240–1247. Waterham, H.R. et al., 2007. A Lethal Defect of Mitochondrial and Peroxisomal Fission. New England Journal of Medicine, 356(17), pp.1736–1741. Weingarten, M.D. et al., 1975. A protein factor essential for microtubule assembly. Proceedings of the National Academy of Sciences, 72(5), pp.1858–1862. Wells, R.C. et al., 2007. Direct Binding of the Dynamin-like GTPase, Dnm1, to Mitochondrial Dynamics Protein Fis1 Is Negatively Regulated by the Fis1 N- terminal Arm. Journal of Biological Chemistry, 282(46), pp.33769–33775. Wild-Bode, C. et al., 1997. Intracellular Generation and Accumulation of Amyloid β-

122

Peptide Terminating at Amino Acid 42. Journal of Biological Chemistry, 272(26), pp.16085–16088. Wu, H.-Y. et al., 2010. Amyloid beta induces the morphological neurodegenerative triad of spine loss, dendritic simplification, and neuritic dystrophies through calcineurin activation. The Journal of neuroscience : the official journal of the Society for Neuroscience, 30(7), pp.2636–49. Wu, J., Anwyl, R. & Rowan, M.J., 1995. beta-Amyloid selectively augments NMDA receptor-mediated synaptic transmission in rat hippocampus. Neuroreport, 6(17), pp.2409–13. Xie, H. et al., 2013. Mitochondrial Alterations near Amyloid Plaques in an Alzheimer’s Disease Mouse Model. The Journal of Neuroscience, 33(43), pp.17042–17051. Yan, J. et al., 2015. Blockage of GSK3β-mediated Drp1 phosphorylation provides neuroprotection in neuronal and mouse models of Alzheimer’s disease. Neurobiology of Aging, 36(1), pp.211–227. Yin, J. et al., 2018. NLRP3 Inflammasome Inhibitor Ameliorates Amyloid Pathology in a Mouse Model of Alzheimer’s Disease. Molecular Neurobiology, 55(3), pp.1977– 1987. Yonashiro, R. et al., 2006. A novel mitochondrial ubiquitin ligase plays a critical role in mitochondrial dynamics. The EMBO Journal, 25(15), pp.3618–3626. Zhao, F. et al., 2017. Mfn2 protects dopaminergic neurons exposed to paraquat both in vitro and in vivo : Implications for idiopathic Parkinson’s disease. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1863(6), pp.1359–1370. Zhao, X.-L. et al., 2010. Expression of beta-amyloid induced age-dependent presynaptic and axonal changes in Drosophila. The Journal of neuroscience : the official journal of the Society for Neuroscience, 30, pp.1512–1522. Zhou, L. et al., 2017. Parkinson’s disease-associated pathogenic VPS35 mutation causes complex I deficits. Biochimica et biophysica acta, 1863(11), pp.2791–2795. Zhu, D. et al., 2006. Phospholipases A2 Mediate Amyloid-beta Peptide-Induced Mitochondrial Dysfunction. Journal of Neuroscience, 26(43), pp.11111–11119. Zhu, X. et al., 2005. Oxidative Imbalance in Alzheimer’s Disease. Molecular Neurobiology, 31(1–3), pp.205–218. Züchner, S. et al., 2004. Mutations in the mitochondrial GTPase mitofusin 2 cause Charcot-Marie-Tooth neuropathy type 2A. Nature Genetics, 36(5), pp.449–451.

123