Hfe Gene Variants and Dysregulation of Cholesterol and Iron Metabolism in Alzheimer Disease and Cancer

The Pennsylvania State University

The Graduate School

College of Medicine

HFE GENE VARIANTS AND DYSREGULATION OF CHOLESTEROL AND IRON METABOLISM IN ALZHEIMER DISEASE AND CANCER

A Dissertation in

Molecular Medicine

Fatima Ghazanfar Ali-Rahmani

Submitted in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

December 2012 The dissertation of Fatima Ali-Rahmani was reviewed and approved* by the following:

James R. Connor

Distinguished Professor of Neurosurgery, Neural & Behavioral Sciences, and Pediatrics

Vice Chair of Neurosurgery

Dissertation Advisor

Chair of Committee

Cara-Lynne Schengrund

Emeritus Professor of Biochemistry and Molecular Biology

Ian Simpson

Professor of Neural & Behavioral Sciences

Patricia S. Grigson

Professor of Neural & Behavioral Sciences

Jong K. Yun

Associate Professor of Pharmacology

Co-chair of Intercollege Graduate Degree Program in Molecular Toxicology

Charles Lang

Distinguished Professor of Cellular and Molecular Physiology, and Surgery

Co-Chair, Intercollege Graduate Degree Program in Molecular Medicine *Signatures are on file in the Graduate School

ii ABSTRACT

Cholesterol and iron are essential for a number of cellular processes. Disruption of both cholesterol and iron metabolism have been independently reported to contribute to several diseases including Alzheimer disease (AD) and cancer. AD is a debilitating disease characterized by progressive neuronal loss and memory impairment. Although the underlying cause(s) of AD are not completely understood, iron accumulation and the resultant oxidative stress are well established contributors to the development of AD. Normal iron uptake is mediated by HFE (high iron, also called hemochromatosis protein). H63D and C282Y are two common variants of the HFE protein that are implicated as disease modifying risk factors for AD and several cancers, respectively. The main objective of this thesis was to determine whether a relationship exists between iron and cholesterol metabolism that could be of relevance for understanding the underlying mechanism of diseases such as AD and cancer. This goal was achieved by studying the effect(s) on cholesterol metabolism of two specific variants of HFE (H63D and C282Y) known to affect iron metabolism.

The H63D variant of HFE, associated with dysregulation of iron homeostasis, has been implicated as a risk factor for Alzheimer disease (AD). Apolipoprotein epsilon 4 (APOE4), a known risk factor for AD, is associated with disruption of cholesterol metabolism. Recent studies showed that individuals with both the APOE4 allele and the H63D variant of HFE had a much greater risk and earlier onset of AD than individuals carrying either the APOE4 allele or H63D variant of HFE alone. Therefore, the hypothesis that H63D-HFE is associated with development of cognitive impairment by disrupting cholesterol metabolism thereby contributing to increased neurodegeneration, and increased memory deficits was investigated. The effect(s) of H63D-HFE on expression of cholesterol was studied using SH-SY5Y human neuroblastoma cells transfected to stably express either wild type (WT-) or H63D-HFE. The ~50% reduction in cholesterol content in cells expressing H63D-HFE was accompanied by a significant decrease in expression of 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMGCoAR), and a significant increase in expression of cholesterol hydroxylase (CYP46A1). Consistent with in vitro studies, H67D-HFE (homologous to human H63D-HFE) knock-in mice,

iii showed a genotype and age dependent decline in brain cholesterol relative to WT-HFE controls and similar changes in expression of proteins regulating cholesterol metabolism. The reduction in cholesterol in brains of H67D-HFE mice was accompanied by a significant decrease in expression of synapse and myelin management proteins, significant increase in caspase-3 expression, and cortical and hippocampal neurodegeneration in the brains of 18 and 24 month old mice. Performance on learning and memory tasks (novel object recognition and Barnes maze) indicated that H67D-HFE mice had poorer cognitive function than WT-HFE mice. Magnetic resonance imaging of 24 month-old mice showed a greater reduction in the volume of brains in H67D-HFE mice. Combined these data indicate that the H63D/H67D variant contributes to the disease process in AD by altering cholesterol metabolism as well as that of iron and that these changes are accompanied by neuronal loss, brain atrophy, and memory impairment which are symptoms of AD. The results also provide a possible explanation for why expression of both APOE4 and H63D-HFE is associated with earlier onset of AD.

Disruptions in the maintenance of iron and cholesterol metabolism have also been implicated in several cancers. Therefore, in the second part of this work, the effects of C282Y-HFE on lipid (cholesterol and sphingolipid) metabolism and its possible relevance to cancer phenotype was also studied. Expression of C282Y-HFE altered transcription of a number of enzymes involved in cholesterol and sphingolipid metabolism. More specifically it was associated with significant changes in the expression of genes involved in cholesterol metabolism were consistent with significant elevation in total cholesterol content of C282Y-HFE cells. Of particular interest was the finding that expression of the enzyme sphingosine kinase 1(SphK) and its pro-survival metabolite, sphingosine-1-phosphate (S1P) was elevated. These effects may contribute to the increased risk of colorectal, breast, prostate, ovarian, and brain tumors reported for carriers of C282Y-HFE and explain why it is considered protective against Alzheimer’s disease while H63D-HFE is a risk factor for neurodegenerative diseases. Changes in the gene expression seen for proteins involved in cholesterol metabolism, indicate that increased uptake of cholesterol is the likely cause for the elevation in cellular cholesterol content seen in C282Y-HFE expressing cells. Since simvastatin is an inhibitor of cholesterol synthesis and has been proposed to be beneficial for the treatment of certain

iv cancers, its effect on the viability of C282Y-HFE cells was ascertained. A higher concentration of simvastatin was needed to decrease survival of cells expressing C282Y- HFE than WT-HFE. Because of the significant increase of SphK in C282Y cells, the effect of a SphK inhibitor (SKI) on cell survival was monitored. Again, WT-HFE cells were more sensitive to this treatment than C282Y-HFE cells, reflecting the drug resistant phenotype of C282Y-HFE cells. Treatment of cells with both simvastatin and SKI revealed a small effect on survival. Collectively, these studies indicated that C282Y-HFE markedly affected both cholesterol and sphingolipid metabolism and support the possibility that inhibition of both cholesterol synthesis/uptake and as well as synthesis of S1P as a potential therapeutic strategy for treatment of cancer in carriers of C282Y-HFE.

Taken together, the studies with both genes variants of HFE provide evidence for the role of HFE in lipid metabolism, in addition to iron management. These data argue for stratification of patients in clinical trials based on HFE genotype to better understand disease mechanism and evaluate therapeutic efficacy of treatments.

v TABLE OF CONTENTS

List of Figures ……………………………………………………………………………ix

List of Tables ………………………………………………………………………….....xi

List of Abbreviations…………………………………………………………………….xii

Acknowledgments...………………………………………………………………...... xvii

Chapter 1: Iron, cholesterol, and Alzheimer’s disease……………………………….1

Introduction ……………………………………………………………………………...18

Genetic Risk factors ……………………………………………………………………..20

Mechanisms implicated in AD…………………………………………………………...21

1. Amyloid Hypothesis……………………………………………………………. 22 2. Tau Pathology…………………………………………………………………... 23 3. Mitochondrial Dysfunction …………………………………………………….. 24 4. Oxidative Stress………………………………………………………………… 25 5. Synaptic Faliure ……………………………………………………………….. 26 6. Metal Toxicity………………………………………………………………….. 27 - Iron ………………………………………………………………………… 27 - Iron in AD ………………………………………………………………….. 28 - Iron, HFE, and AD …………………………………………………………. 30 - HFE gene …………………………………………………………………... 31 - Structure of HFE protein……………………………………………………. 32 - Function of HFE ………………………………………………………….. 33 - HFE expression …………………………………………………...………...36 - H63D-HFE and AD………………………………………………………… 36 7. Lipid dyshomeostasis ……………………………………………………………38 Lipid rafts ………………………………………………………………………..38

i) Sphingolipids ………………………………………………………..41

vi -Role of sphingolipids in AD…………………………………… 42

ii) Cholesterol …………………………………………………………..44

- Cholesterol metabolism and function in circulation …………...44 - Cholesterol metabolism in the brain …………………………...45 - Brain vs. plasma cholesterol …………………………………...49 - Role of cholesterol in the brain………………………………... 49

- Cholesterol homeostasis and AD pathology …………………...51

- Cholesterol in aging brain ……………………………………...51

- ApoE functions and isoform-specific effects relevant to AD pathology ………………………………………………………………………………...53

Rationale for thesis ……………………………………………………………………...55

References ………………………………………………………………………………58

Chapter 2: Iron, cholesterol, and Alzheimer’s disease…………………………...…101

Introduction …………………………………………………………………………….101

Experimental Procedures……………………………………………………………… 103

Results ………………………………………………………………………………….111

Discussion ……………………………………………………………………………...121

Figure and legends…………………………………………………………………….. 130

References ……………………………………………………………………………...143

Chapter 3: Effect of HFE variants on sphingolipid expression by SH-SY5Y human neuroblastoma cells…………………………………………………………………... 155

Experimental Procedures……………………………………………………………….157

Results ………………………………………………………………………………….160

vii Discussion……………………………………………………………………………... 162

Figures…………………………………………………………………………………. 165

References ……………………………………………………………………………...173

Chapter 4: C282Y increases cellular proliferation by inducing cholesterol accumulation…………………………………………………………………………..178

Introduction ……………………………………………………………………………179

Role of iron in cancer …………………………………………………………………..179

C282Y in cancer ……………………………………………………………………….181

Role of cholesterol in cancer …………………………………………………………..182

Methods and materials …………………………………………………………………184

Results ………………………………………………………………………………….185

Discussion ……………………………………………………………………………...188

Tables and figures…………………………………………………………………… 193

References …………………………………………………………………………….199

Chapter 5 Summary and Future perspectives……………………………………. 212

Introduction …………………………………………………………………………..212

Is there a link between iron and cholesterol metabolism? ……………………………215

How does H63D-HFE affect interactions between glial and neuronal cells? ………...219 Potential biomarkers for AD progression ……………………………………………..222

Is it iron accumulation or expression of the HFE variants that is the culprit? ………...224

Therapeutic implications: Is statin therapy appropriate for AD? ……………………...226

References ……………………………………………………………………………..228

viii Appendix: Cholesterol, GM1, and Autism

Introduction ……………………………………………………………………………236

Methods………………………………………………………………………………... 238

Results ………………………………………………………………………………….239

Discussion……………………………………………………………………………... 240

Figures ………………………………………………………………………………….244

References……………………………………………………………………………. 247

ix List of Figures

Figure 1.1: Role of HFE and its variants in iron regulation ……………………………..35

Figure 1.2: Cholesterol synthesis pathway……………………………………………... 45

Figure 1.3: Mechanism of cholesterol transport between astrocytes and neurons in the brain ……………………………………………………………………………………..48

Figure 1.4: Schematic to show redistribution of membrane lipids to allow cholesterol efflux via 24S-HC………………………………………………………………………. 48

Figure 2.1: Effect of expression of H63D-HFE on cholesterol and cholesterol metabolizing enzymes …..…………………………………………………………...... 132

Figure 2.2: Effect of cholesterol manipulation on survival of HFE-variant cells……………………………………………………………………………………. 133

Figure 2.3: Effect of expression of H67D-HFE on expression of cholesterol and cholesterol metabolizing enzymes in brains of mice ……………………….………….134

Figure 2.4: H67D-HFE alters expression of synaptic proteins ………………………...135

Figure 2.5: Effect of H67D-HFE on expression of proteins involved in myelin maintenance ………………………………………………………………………….136

Figure 2.6: H67D-HFE induces neurodegeneration..………………………………...137

Figure 2.7: Total brain volume (gray and white matter) of 22 month old WT- and H67D- HFE expressing mice………………………………………………………...... 138

Figure 2.8: Effect of expression of H67D-HFE on learning and memory-

Latency measures……………………………………………………………………... 139

Figure 2.9: Effect of expression of H67D-HFE on learning and memory-

Total Distance & errors…………………………………………………………………140

x Figure 2.10: Correlation analysis ………………………………………………………141

Figure 2.11: Serum cholesterol content of WT- and H63D-HFE mice……………….. 142

Figure 3.1: Cell morphology and growth curves ………………………………………166 Figure 3.2: Association of FLAG-tagged WT-HFE with lipid rafts …………………...167 Figure 3.3: HPTLC analysis of the ganglioside composition of cells transfected with either WT, H63D, or C282Y variants of HFE……………...... 168 Figure 3.4: Binding of Alexa-Fluor®594-conjugated CTxB to the cell surface of SH- SY5Y human neuroblastoma cells stably transfected with either WT-, H63D-, or C282Y- HFE……………………………………………………………………………………..169 Figure 3.5: Steps in sphingolipid metabolism catalyzed by the enzymes studied……………………………………………………………………………...…...170 Figure 3.6: Expression of sphingosine kinase 1………………………………………...171 Figure 3.7: S1P synthesized by cell lysate SphK1…………………………………….. 172 Figure 4.1: Effect of expression of C282Y-HFE on cholesterol content relative to WT- HFE expressing cells…………………………………………………………………... 194

Figure 4.2: Effect of simvastatin treatment on survival of WT- and C282Y-HFE cells…………………………………………………………………………………..…195

Figure 4.3 Effect of sphingosine kinase inhibitor (SKI) on cell survival A) SKI alone 24 hr B) SKI alone 48 hr………………………………………………………………... 196

Figure 4.4: Effect of combination treatment of simvastatin and sphingosine kinase inhibitor on cell survival treatment- 10uM of statin………………………………….. 197 Figure 4.5: Effect of combination treatment of simvastatin and sphingosine kinase inhibitor on cell survival treatment- 40uM of statin………………………………….. 198 Figure 5.1: Summary of findings from this thesis ……………………………………..214

Figure 6.1: Cholesterol content of RBC membrane……………………………………………………………………………… 244 Figure 6.2: GM1 expression in the RBC membranes…………………………………………………………………………….. 245

xi List of Tables

Table 2.1: Expression of genes involved in cholesterol metabolism by H63D-HFE- expressing cells relative to WT-HFE-expressing cells ……………………………….130

Table 2.2: Summary of cellular, biochemical, anatomical, and behavioral findings in H63D/ H67D-HFE relative to WT-HFE cells and mice………………………………. 131

Table 3.1: Relative expression of mRNAs for 14 enzymes involved in sphingolipid

Synthesis……………………………………………………………………………… 165

Table 4.1: Effect of C282Y-HFE variant on the expression of genes involved in cholesterol metabolism relative to WT-HFE-expressing cells……………………….. 193

Table 5.1: Comparison of in vivo and in vitro findings of H63D- and C282Y-HF…….225 Table 6.1: Correlation plot between GM1 (intensity of CTxB) versus cholesterol content of RBC membrane from autistic children ………...……………………………………246

xii LIST OF ABBREVIATIONS

24S-HC 24S-hydroxycholesterol aa amino acid

Aβ amyloid‐beta

ABCA7 ATP-binding cassette transporter

ACAT acyl-coenzyme A cholesterol acyltransferase

Ach Acetylcholine

AD Alzheimer’s disease

AKT protein kinase B

ALS amyotrophic lateral sclerosis

ANOVA analysis of variance

APOD apolipoprotein D

ApoE Apolipoprotein E

ApoE4 Apolipoprotein E4

APP amyloid precursor protein

ATP adenosine triphosphate

B1N1 bridging integrator 1

BACE β‐amyloid converting enzyme

BBB blood‐brain barrier

BDNF brain-derived neurotrophic factor

xiii CD2AP CD2-associated protein

CD33 sialic acid-binding immunoglobulin-like lectin

Cdk cyclin‐dependant kinase

CDK5 cyclin-dependant kinase 5

CLU Clusterin

CNS central nervous system

COLEC12 collectin sub-family member 12

CR1 complement receptor 1

CSF cerebrospinal fluid

CYP46A1 cholesterol hydroxylase

CYP7B1 cytochrome P450, family 7, subfamily B, polypeptide 1

DFO desferrioxamine

DMEM Dulbecco’s modified Eagle’s medium

DMT divalent metal transporter

DNA deoxyribonucleic acid

DRM detergent-resistent microdomains

DRP1 dynamin-related protein 1

DS Down syndrome

EOAD Early-onset Alzheimer’s disease

EPHA1 ephrin receptor A1

ER endoplasmic reticulum

xiv FAC ferric ammonium citrate

FAD familial Alzheimer’s disease

FDA Food and Drug Administration

Fe2+ ferrous iron

Fe3+ ferric iron

GFAP glial fibrillary acidic protein

GI gastrointestinal

GM‐CSF granulocyte‐macrophage colony stimulating factor

GPI glycosylphosphatidylinositol

GSK glycogen synthase kinase

GSK-3β glycogen synthase kinase 3β

GWAS Genome-Wide Association Studies

HDL high density lipoproteins

H2O2 hydrogen peroxide

HFE protein product of HFE gene

HH hereditary hemochromatosis

HHC hepatocellular carcinoma

HMGCoAR 3-hydroxy-3-methylglutarylcoenzyme A reductase

HRP horse radish peroxidase

IDL intermediate-density lipoproteins

IL interleukin

xv IRE iron responsive element

IRP iron regulatory protein kD kilodalton

LDL low density lipoproteins

LDLR low density lipoprotein receptor

LIP labile iron pool

LOAD late-onset Alzheimer’s disease

LPS lipopolysaccharides

LRP1 LDLR-related protein

LRP1B low density lipoprotein-related protein 1B

LTP long-term potentiation

LTD long-term depression

M‐CSF macrophage colony stimulating factor

MβCD methyl-ß-cyclodextrin

MCI mild cognitive impairment

MDM2 murine double minute 2

MHC major histocompatibility complex

MPT mitochondrial permeability transition pore mRNA messenger ribonucleic acid

MRI magnetic resonance imaging

MS multiple sclerosis

xvi MS4A membrane-spanning 4-domains subfamily A

MTT [3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyl tetrazolium bromide]

NeuN neuron‐specific nuclear protein

NFT neurofibrillary tangles

NMDA N‐methyl D‐aspartate

NPC1L1 Niemann-Pick C1-Like 1

O2 oxygen

OH∙ hydroxyl radical

ORF Open reading frame

PBS phosphate-buffered saline

PD Parkinson’s disease

PDH pyruvate dehydrogenase

PEN‐2 presenilin enhancer‐2

PICALM phosphatidylinositol-binding clathrin assembly protein

PKB Akt/Protein Kinase B

PS‐1/2 presenelin‐1 or presenilin‐2

PVDF polyvinylidene fluoride

RIPA radioimmunoprecipation assay

RBC Red blood cells

ROS reactive oxygen species sAPP soluble amyloid precursor protein

xvii S1P Sphingosine-1-phosphate

SKI SphK inhibitor

SphK1 sphingosine kinase 1

SLOS Smith-Lemli-Opitz syndrome

SNP single nucleotide polymorphisms

SH‐SY5Y human neuroblastoma cells

TAG triacylglycerides

TCA tricarboxylic acid

Tf transferrin

TfR transferrin receptor

TLR toll‐like receptor

TNF tumor necrosis factor

VLDL very low density lipoproteins

VDAC voltage‐dependant anion channel

WT wild type

ADI-R Autism Diagnostic Interview-Revised

M-CHAT Modified Checklist for Autism in Toddlers

CTxB cholera toxin

Frt Ferritin

P16INK4a cyclin-dependent kinase inhibitor

xviii Acknowledgements

This thesis is dedicated to my parents and siblings who have supported, loved, and encouraged me throughout my life. My gratitude for you cannot be expressed in words. I appreciate everything you have done for me specially keeping up with me during stressful times. You served as my backbone to get me through those rough times. I am lucky to have a family like you.

My heartfelt appreciation for my co-advisors Dr. Cara-Lynne Schengrund and Dr. James Connor who have cooperated and worked together and provided exceptional guidance on my thesis project throughout my graduate work and also helped me develop as an independent scientist. To Dr. Schengrund, I really appreciate you staying for an extra year despite your plans to retire and especially for continuing your critical guidance after your retirement. To Dr. Connor, I truly appreciate your critical feedback on grant writing and the freedom to pursue research questions, as well as, the opportunities you have given me to attend professional conferences for my professional development. There were several moments when I was frustrated and not so positive and at those times both of you have raised my morale with positive comments, I would always treasure those moments. I am very lucky to have the two best advisors who trained me like their own child with respect and diligence.

I sincerely would like to thank members of Biochemistry department, as well as, all current and past members of ‘Connor lab’. You all have been extremely supportive to me. I especially thank Dr. Sang Lee, Dr. Madhan, Dr. Padma, Dr. Amanda Snyder, Beth Neely, and Wint Nandar for your tremendous guidance and support throughout my work in lab. At many times you were my life savers.

I thank all my committee members, people at the genomic, confocal and animal core facilities, staff at the graduate office for their help and support. I am also thankful of security officers who have offered me a ride home on all those nights I was working late in lab. I also thank all my friends- Dr. Becca Weinberg, Dr. Bozho Todorich, Dr. Markus Ackermann, Dr. Nida Rizvi, Dr. Janet, Dena, Amber, Su-Fern, Zeinab, Lynnsay, Bill and Wint- whose support and encouragement has kept me going during my graduate work.

xix Bismillah hirrahmanirrahim

Chapter 1

Iron, Cholesterol, and Alzheimer’s disease

Introduction

Disease and Prevalence

Alzheimer’s disease (AD), first described over 100 years ago by Alois Alzheimer (Berchtold and Cotman, 1998) is a neurodegenerative disorder characterized by memory loss, spatial disorientation, and reduced mental abilities. AD is the most common cause of dementia (50-60% of dementia cases) and the 6th leading cause of death. In the United States, more than 5.3 million people are living with AD [alz.org]. According to the current estimates, women (16%) over 71 years old are at a higher risk for developing AD than men (11%). With the population of people aged 65 and over expected to increase from the ~12% that it was in 2006 to ~20% in 2030. It is clear that if nothing is done, the problem will continue to grow and become a greater burden on the health care system. Though increasing age is the strongest risk factor of AD with the highest prevalence of the disease in individuals age 65 or over, a subset of younger population is also affected with AD.

AD is generally categorized as either early-onset or late-onset based on age of disease onset. Classification of AD is illustrated in the table below. Early-onset AD (EOAD) accounts for about 5% of the 5.4 million people who experience disease symptoms and those symptoms can start to appear in people as young as their 30s. However, late-onset AD (LOAD) cases, symptoms start to appear in their 60s or later, constitute 95% of population affected by AD. Both types of AD have a significant

1 genetic component, with an estimated heritability of 5-25 % for EOAD and 58–79% for LOAD (Gatz et al., 2006)(Wingo et al., 2012).

Pathogenesis

The characteristic pathological features of AD are the deposition of extracellular senile amyloid plaques and formation of intracellular neurofibrillary tangles (NFTs). Accumulation of amyloid β (Aβ) deposits and formation of NFTs (due to hyper phosphorylation of tau protein) have long been believed to be underlying pathological events disrupting synaptic communication and promoting neuronal death (Alzheimer et al., 1995; Selkoe, 2002a; Terry, 1963; Walsh and Selkoe, 2004).

In AD, the areas of the brain that are particularly vulnerable to synaptic loss, neuronal loss and atrophy are those involved in memory formation and cognition such as the limbic and association cortices and some subcortical nuclei with large cortical projections (LaFerla; Walsh and Selkoe, 2004). Imaging and autopsy studies show atrophy of the hippocampal formation, memory center of the brain, and of the cerebral cortices along with enlarged ventricles in the brains of AD patients (Perl, 2010; Walsh and Selkoe, 2004). The entorhinal cortex is the region first affected, followed by degeneration extending to frontal and temporal cortices (Walsh and Selkoe, 2004). Clinical stages of disease

AD progression is categorized into six stages: stages I–II represent the clinically silent involvement of transentorhinal cortex; stages III–IV are characterized by lesions in the entorhinal/transentorhinal regions and correspond to the phase of mild cognitive decline; in stages V–VI, there is severe neocortical destruction and fully developed dementia (Braak and Braak, 1995; Perl, 2010). Mild cognitive impairment (MCI) is considered to be a transitional stage between normal aging and AD (Grundman et al., 2004; Petersen et al., 1999).

Diagnosis

Current diagnosis is based on severity of the clinical symptoms of memory loss which is usually only detected in an advanced stage of the disease. Confirmed diagnosis of AD

2 usually comes from autopsy studies showing existence of plaques and tangles or brain atrophy. Recent advances using magnetic resonance imaging (MRI) technology have allowed visualization of amyloid plaques in the brain but those are not always associated with cognitive decline.

Treatment

Despite decades of research there is no cure available for AD. After disease onset, symptoms are usually progressive. Currently available treatments only manage and slow progression of symptoms, they do not treat the underlying pathology. Because early research in patients with AD showed reduction in synthesis of the neurotransmitter acetylcholine, current therapy options are focused on inhibiting the enzyme responsible for breakdown of acetylcholine (cholinesterase). Currently approved drugs are cholinesterase inhibitors: donepezil (Aricept ®), galantamine (Razadyne ®), and rivastigmine (Exelon ®), and a glutamate inhibitor: memantine (Namenda ®) (Shah et al., 2008). These drugs are associated with significant side effects that limit their use. Despite a heavy focus of scientific research in the past few decades on markers of amyloid and tau pathology, none of the drugs targeted for amyloid and Tau are currently approved by the Food and Drug Administration (FDA) for clinical use. Several drugs targeted to reduce amyloid plaques are being tested in clinical trials. Unfortunately, a recent report showed two of these drugs, bapineuzumab and solanezumab, failed to show any benefit in two large trials (Callaway, 2012), leading to questioning of the relevance of the amyloid hypothesis in the etiology of AD. Since, hypercholesterolemia is considered to be a risk factor of AD, use of statins (class of drugs that inhibit cholesterol synthesis) has been assessed but these studies have given conflicting results. Some have reported memory improvement with use of statins while others have shown either no therapeutic or preventive effects or detrimental effects on memory function. Because AD is a multifactorial disease with evidence for multiple causes and genetic mutations in genes with diverse function, it would be beneficial to identify subpopulations responsive to a certain treatment based on their physiology. Identification of genetic and environmental factors and knowledge about their interactions and resulting pathological markers will allow us to stratify AD patients for improved therapeutic outcomes.

3 Genetic risk factors:

Inheritance of mutant forms of a number of identified genes encoding proteins such as amyloid precursor protein (APP), Presenilin 1 (PS1) and/ or Presenilin 2 (PS2) have been attributed to EOAD or familial AD (Goate et al., 1991; Selkoe, 1996; Thinakaran, 1999). Apolipoprotein E4 (ApoE4) is the strongest candidate in conferring risk for late onset AD (LOAD) and this association was first described in the early 1990s (Corder et al., 1993; Saunders et al., 1993; Strittmatter et al., 1993a; Strittmatter et al., 1993b). Since then the APOE gene has been repeatedly implicated in AD via linkage studies and these findings have been confirmed by Genome-Wide Association Studies (GWAS) (Abraham et al., 2008; Bertram et al., 2008; Carrasquillo et al., 2009; Coon et al., 2007; Grupe et al., 2007; Heinzen et al., 2012; Li et al., 2008; Potkin et al., 2009; Reiman et al., 2007). The contribution of allelic isoforms of APOE to AD will be discussed later. Recently, four large LOAD GWAS (Harold et al., 2009; Hollingworth et al., 2011a; Hollingworth et al., 2011b; Lambert et al., 2009; Naj et al., 2011; Seshadri et al., 2010) have replicated association of APOE to AD and have also identified nine novel loci, collectively representing 50% of LOAD genetics. These genes and the proteins they encode are CLU– clusterin, PICALM– phosphatidylinositol-binding clathrin assembly protein, CR1– complement receptor 1, BIN1– bridging integrator 1, ABCA7– ATP-binding cassette transporter, MS4A cluster – membrane-spanning 4-domains subfamily A, CD2AP– CD2- associated protein, CD33– sialic acid-binding immunoglobulin-like lectin and EPHA1– ephrin receptor A1. Based on the known functions of these genes, the importance of pathways involving function of the immune system, cholesterol metabolism and synaptic cell membrane processes in AD has been heightened (Morgan, 2011). While association of these genes with AD does not strongly implicate the amyloid hypothesis in AD etiology, it is possible that they play indirect roles in promoting amyloid. Regardless, these studies provide paradigm shifting evidence and open up avenues for scientific research to look at additional pathways in the etiology of AD. The one selected for study in this thesis was whether there is an association between alterations in cholesterol metabolism and onset of dementia.

4 Mechanisms implicated in AD

A number of possible mechanisms leading to the neurodegeneration seen in AD have been proposed including 1) the amyloid hypothesis (Morishima-Kawashima and Ihara, 2002), Tau hyperphosphorylation (Yankner, 1996; Yankner et al., 2008), 2) mitochondrial dysfunction (Reddy, 2011), 3) oxidative stress (Cai et al., 2011), 4) synaptic dysfunction (Arendt, 2009), 5) metal toxicity (Bush, 2003; Connor and Lee, 2006; Connor et al., 2001a; Liu et al., 2006; Maynard et al., 2005; Roberts et al.; Sayre et al., 2000) and 6) disruption of lipid metabolism (Alessenko et al., 2004; Canevari and Clark, 2007; Farooqui et al., 2007; Hartmann et al., 2007; Koudinov et al., 2002; Koudinov and Koudinova, 2003; Koudinov and Koudinova, 2005; Lukiw, 2006). These pathways will be briefly discussed below with more in depth discussion of the roles of iron and of lipid metabolism in AD. It is not clear whether disruption of one of these pathways is the initiating event in the pathogenesis of AD; it is possible that multiple pathways are interconnected and a disruption in one could affect another, triggering a chain reaction and worsening of phenotype.

1. Amyloid hypothesis

Proteolytic processing of amyloid precursor protein (APP) gives rise to the β-amyloid protein (Aβ). This processing is catalyzed by β and γ secretases (Glenner and Wong,

1984) and results in Aβ forms of varying length such as Aβ1–40 and Aβ1–42. Unless favored by mutations in the APP gene, or genes encoding secretase complex components, the latter is less prevalent but more hydrophobic. As a consequence of the increased hydrophobicity, it is more likely to aggregate than the Aβ1–40 isoform (Perl, 2010) forming amorphous aggregates, soluble oligomers or amyloid fibrils. For a long time, plaques of insoluble fibrillar Aβ were believed to initiate AD pathogenesis (“amyloid cascade hypothesis”). This notion, however, was not corroborated by neuropathological studies because the degree of dementia was not found to be commensurate with the amount of amyloid plaques (Ingelsson et al., 2004; Nagy et al., 1996; Terry et al., 1991).

5 According to current understanding, oligomeric aggregates exert cytotoxic effects by exposing “promiscuous hydrophobic surfaces” (Hartl et al., 2011). These surfaces are believed to capture metastable proteins for co-aggregation, thus depleting those that might function as transcription factors, regulators of translation or as key components of the architecture of a cell. This scenario is exemplified by the early onset familiar forms of AD where intraneuronal Aβ oligomers are demonstrably at the origin of a neurotoxic cascade (Gotz et al., 2004). This cascade includes mitochondrial dysfunction (Amadoro et al., 2012; LaFerla et al., 2007), formation of reactive oxygen species, hyperphosphorylation of tau protein (LaFerla and Oddo, 2005; Walsh and Selkoe, 2004), microgliosis and astrocytosis (Walsh and Selkoe, 2004), impairment of synaptic function and neurotransmission (Bao et al., 2012; Walsh and Selkoe, 2004). Eventually, this leads to the known cognitive decline (Arendt, 2009; LaFerla and Oddo, 2005; Walsh and Selkoe, 2004).

In contrast, what triggers neurotoxicity in the more prevalent, sporadic form of AD is still a matter of debate (Gotz et al., 2004). In the case of early onset forms of AD, mutations in APP (De Jonghe et al., 1998; Haass et al., 1995; Mullan et al., 1992; Murrell et al., 2000; Nilsberth et al., 2001; Suzuki et al., 1994) or β and γ secretases cause overproduction of Aβ and its aggregation (Selkoe, 2005). It is conceivable that in late- onset AD, different mechanisms are at work, albeit resulting in the same trigger, oligmeric Aβ. Current knowledge suggests that the collapse of proteostasis in aging individuals indirectly promotes aggregation due to an impairment of the chaperone system that prevents aggregation under normal conditions (Gotz et al., 2004). Moreover, clearance of aggregates by autophagy is also impaired in the ageing process (Hartl et al., 2011).

2. Tau pathology

On the histological level, AD brains display two characteritics: amyloid plaques and neurofibrillary lesions. Both are caused by aggregation of proteins. Neurofibrillary tangles are formed by the hyperphosphorylated form of a microtubule-associated protein called tau and several proteins that co-aggregate, e.g. ubiquitin (Perry et al., 1987),

6 cholinesterases (Mesulam and Moran, 1987) and A4 amyloid protein (Hyman et al., 1989). Tau mediates assembly and stability of microtubules in axons and thus plays a crucial role in axoplasmatic transport. Tau phosphorylation regulates its affinity for microtubules (Perl, 2010). In line with this, the abnormal pattern of tau hyperphosphorylation observed in AD causes detachment from axonal microtubules and promotes protein aggregation forming NFTs. As mentioned above, the dysfunction of tau affects cellular trafficking, in particular the delivery of mitochondria to axons (Reddy, 2011).

There is evidence to suggest that tau aggregation and the resulting neurofibrillary degeneration occur upstream of Aβ aggregation. Indeed, it was proposed to initiate a multifactorial, pathogenic response including oxidative damage and glial activation, mitochondrial dysfunction and neuronal damage (Gotz et al., 2004). On the other side, upstream events of tau aggregation are just starting to be understood. Several protein kinases have been reported to share in tau hyperphosphorylation. These include cyclin dependent kinase 5 (CDK5), glycogen synthase kinase 3β (GSK3β), and extracellular signal-related kinase 2 (ERK2; (Ballard et al., 2011)).

3. Mitochondrial dysfunction

AD is characterized by pathological calcium fluxes resulting in calcium overload (Querfurth and LaFerla, 2010). To understand how calcium affects mitochondrial function in AD, it is necessary to review some basic concepts.

Mitochondria are important sites of calcium storage inside the cell where they are physically tethered to the endoplasmic reticulum by means of interactions between mitofusin proteins (Decuypere et al., 2010). This tethering creates a microdomain where activation of IP(3) receptors results in elevated calcium concentrations(Decuypere et al., 2010). Calcium is subsequently taken up by the mitochondria in order to stimulate ATP production. Alternatively, calcium signals can be propagated within the mitochondrial network, thus shaping the spatio-temporal characteristics of calcium-dependent signaling cascades (Decuypere et al., 2010). In order to do so, mitochondria have to maintain fused mitochondrial networks. Mitochondrial morphology is shaped by ongoing fusion and

7 division events. Intriguingly, inhibition of mitochondrial division leads to an attenuation of pathological events in cell culture models of AD (Barsoum et al., 2006; Cho et al., 2009). At the molecular level, this results in inhibition of dynamin-related protein 1 (Drp1), a GTPase that mediates mitochondrial division (Lackner and Nunnari, 2010).

Besides this “healthy” calcium handling, mitochondria can be adversely and severely affected by a calcium overload. First, calcium activates calcineurin which in turn dephosphorylates and activates Drp1 leading to mitochondrial fragmentation and, ultimately, to apoptosis(Oettinghaus et al., 2012). Second, an excess of calcium disrupts the integrity of mitochondrial protein complexes. For example, calcium sequesters cardiolipin, thus depriving the ADP/ATP carrier of a stabilizing and activating protein- lipid interaction. Not only is the ADP/ATP carrier vital, it also associates with respiratory complexes, enabling the formation of supercomplexes (Klingenberg, 2009). An impairment in formation of respiratory supercomplexes results in production of reactive oxygen species which enhance aggregation of tau.

4. Oxidative stress

Numerous studies have implicated oxidative stress as a contributing factor in AD (Alessenko et al., 2004; Cai et al., 2011; Castellani et al., 2012; Cutler et al., 2004a; Markesbery, 1997; Perry et al., 2003; Perry et al., 2000; Rottkamp et al., 2000; Uttara et al., 2009; Yankner et al., 2008). Excessive production of reactive hydroxyl radicals that when outcompetes cellular antioxidant mechanisms culminate in oxidative stress. In fact, mitochondrial dysfunction in AD can lead to the generation of reactive oxygen species (ROS). This is, at least partially, triggered by excessive calcium. Being a divalent cation, calcium is able to sequester the lipid dimer cardiolipin, thus depriving respiratory complexes and mitochondrial carriers of stabilizing and activating interactions (Klingenberg, 2009).

However, respiratory complexes are not the only source of reactive oxygen species. Iron potentially serves as a major source for hydroxyl radicals as exemplified by the so-called Fenton reaction (Lloyd et al., 1997).

3+ 2+ Fe + •O2 → Fe + O2 (Step I)

2+ 3+ - Fe + H2O2 → Fe + OH + •OH (Step II)

For this reason, iron requires constant “chaperoning” (Richardson et al., 2010) to prevent toxic effects mediated by ROS. These effects include covalent modifications of nucleic acids, proteins, and lipids such as oxidation of amino groups, crosslinking, and peroxidation, respectively (Uttara et al., 2009). Importantly, these modifications are particularly detrimental for cells of the immune and nervous systems Free radicals, lipid peroxidation and antioxidants in apoptosis: implications in cancer, cardiovascular and neurological diseases (Ferrari, 2000). Indeed, ROS have been implicated in neuronal cell death. For instance, ROS impair lipid metabolism in such a way that ceramides accumulate which in turn trigger toxic and pro-apoptotic effects (Cutler et al., 2004a). Interestingly, ROS were reported to abrogate normal calcium signaling in the context of neuronal cell death (Ermak and Davies, 2002). Thus, iron overload triggers ROS production and ROS will reinforce dysregulation of calcium signals, creating a vicious circle-like feedback loop. In summary, ROS establish a pathogenic burden by means of unspecific oxidation reactions that are able to target virtually all molecules in a cell including proteins, RNA, DNA, and lipids.

The need to chaperone iron increases during the ageing process because iron – just like other redox-active metal ions – accumulates in the brains of ageing individuals (Takahashi et al., 2001). This showed to be important for the development of AD as Aβ was shown to bind iron, thereby initiating the Fenton Reaction (Opazo et al., 2002). Still, it has been a matter of debate which oligomeric state of Aβ represents the toxic species. The current opinion holds that soluble oligomers which are less structured than amyloid fibrils are the primary toxic species (see section “Amyloid hypothesis”). Since these oligomers expose “promiscuous hydrophobic surfaces” (Hartl et al., 2011), they are able to unspecifically bind a multitude of proteins and other molecules that will be subject to

9 Fenton-like oxidation processes catalyzed by iron. Of note, certain genetic predispositions like those causing hemochromatosis favor the accumulation of iron (Connor and Lee, 2006; Connor et al., 2001a; Connor et al., 1992a; Nandar and Connor) and will be discussed below in detail.

5. Synaptic failure

The mental decline observed in AD has been consistently correlated with the physical loss of synapses (Arendt, 2009; Coleman and Yao, 2003; Koudinov and Koudinova, 2001; Koudinov and Koudinova, 2003; Masliah et al., 2001; Querfurth and LaFerla, 2010; Selkoe, 2002b; Sze et al., 1997; Terry et al., 1991; Walsh and Selkoe, 2004). Several studies indicated that the brains of AD patients have reduced expression of a number of proteins involved in synaptic vesicle trafficking, with concomitant depletion of neurotransmitter, altered synaptic transmission, and synaptic/neuronal loss (Arendt, 2009; Coleman and Yao, 2003; Yao et al., 2003). Aβ oligomers adversely affect synaptic plasticity such that long-term potentiation (LTP) and long-term depression (LTD) are dysregulated (LaFerla and Oddo, 2005). Indeed, high levels of oligomers are thought to undermine synaptic transmission. For instance, Aβ was shown to promote endocytosis of NMDA and AMPA receptors (LaFerla and Oddo, 2005). Postsynaptic membranes devoid of these receptors will no longer be able to trigger LTP. On the other hand, Aβ is able to induce calcium influx through NMDA receptors which leads to elevated intracellular calcium concentrations and subsequent activation of aberrant signaling cascades which facilitate LTD. Moreover, Aβ interacts with the receptors of acetylcholine (Ach) and brain-derived neurotrophic factor (BDNF), thus abrogating Ach signaling and adding to the depletion of BDNF, respectively (Querfurth and LaFerla, 2010). Dendritic spines are severely affected in AD because tau hyperphosphorylation impairs maintenance of these microcompartments by abrogating cytoskeletal transport (Adalbert et al., 2007; Reddy, 2011).

6. Metal toxicity

Numerous studies have implicated metals such as iron, copper, zinc and aluminum in the pathogenesis of AD (Bush, 2003; Connor and Lee, 2006; Connor et al., 2001a; Liu et al.,

10 2006; Maynard et al., 2005; Roberts et al.; Sayre et al., 2000). Much evidence supports the role of iron in AD is thought to be mediated by resulting oxidative stress as described above. Because iron is the most abundant transition metal in the body and is readily available from several dietary options, it is important to understand how its regulation in the body is influenced by other factors and how its elevation or depletion affects cellular processes that could lead to AD pathogenesis. To understand the role of iron in neurodegeneration, I will first describe its normal function and regulation.

Iron

Iron is essential for a number of cellular processes needed for survival. Iron is a required cofactor for a number of enzymes involved in cellular energy production (mitochondrial electron transport chain) (Muhlenhoff et al., 2010), DNA synthesis and repair, ribosome biogenesis, neurotransmitter synthesis, myelin synthesis and lipid metabolism, and cell cycle regulation. Iron is also needed for heme production and formation of iron-sulfur clusters (Rouault and Tong, 2005; Tong and Rouault, 2006). Though these reactions occur in every organ, the role of iron in the brain is particularly important. Brain has the highest demand for oxygen consumption of all organs and iron-containing hemoglobin is essential to meet this high need of oxygen demand. In addition, to maintain the complex communication network and to establish plasticity, the brain constantly remodels cell contacts and synapses. These processes rely on protein synthesis by ribosomes which, in turn, depend on [4Fe–4S] cluster-containing proteins (Kispal et al., 2005). These iron- sulfur clusters are also essential for DNA synthesis (Netz et al., 2012). Moreover, iron is an essential trophic factor that is necessary for myelination. This is because key enzymes such as HMGCoA reductase, squalene synthase, glucose-6-phosphate dehydrogenase needed for the biosynthesis of myelin cholesterol and lipids (e.g. 25-30% cerebrosides) require iron as a cofactor (Lange and Que, 1998; Todorich and Connor, 2004). Role of iron in such wide range of processes is thought to be due to its ability to 1) donate electrons needed for redox reactions catalysed by many enzymes, 2) transfer electrons in mitochondria, and 3) to bind oxygen in heme. But if iron accumulates in an unchelated form, this same ability of iron to readily exchange electrons can result in formation of reactive free radicals by the Fenton reaction (Lloyd et al., 1997). In turn, this can lead to

11 oxidative stress which can negatively impact a number of cellular processes and disrupt cell membrane integrity. Because oxidative stress is a leading hypothesis in the etiology of AD, the role of iron and disorders resulting in iron overload have been under intense investigation for their potential role in the risk for AD.

Iron in AD

Several studies have reported that brain iron content and expression of iron-regulating proteins such as Ferritin (Frt; high expression indicates more iron) increase with age, with a much greater increase in iron content than the protein that stores it i.e. Frt (Bartzokis et al., 1994; Bartzokis et al., 2007; Connor et al., 1992a; Connor et al., 1995; Connor et al., 1992b; Hallgren and Sourander, 1958; Hirose et al., 2003; Martin et al., 1998; Milton et al., 1991). The increase in iron has been proposed to underlie age-related cognitive decline (Bartzokis, 2009). However, there is substantial evidence that iron accumulation found in the brains of AD and patients with mild cognitive impairment (MCI) is significantly higher than in those of age-matched non-demented controls (Ding et al., 2009). Disturbances in iron metabolism have been found in post-mortem brain tissue, in cerebrospinal fluid (CSF) and in vivo using magnetic resonance imaging (MRI). Regions where iron deposition has been found include the hippocampus, basal ganglia and cortex (Bartzokis et al., 2000; Ding et al., 2009; Loeffler et al., 1995) as well as in senile plaques and neurofibrillary tangles and cells surrounding them (Connor et al., 1992a). Moreover, higher brain iron content has also been reported to correlate with the severity of cognitive impairment. Thus, brain iron evaluation by MRI has been proposed as a putative marker for assessing disease progression. Excess unchelated iron has been considered a major cause of oxidative stress that can lead to modification of proteins, lipids, DNA, and RNA, thereby, inducing several features associated with AD (Ferrari, 2000; Markesbery, 1997). As indicated above, these alterations are toxic to cells because they result in activation of cell apoptosis pathways and eventually cell death. In fact, markers of protein oxidation and lipid peroxidation have been consistently found elevated in the brain and CSF samples of AD and MCI patients who subsequently developed AD (Brys et al., 2009; Markesbery, 1997; Pratico et al., 2002). Moreover, markers of lipid peroxidation in ventricular fluid have been correlated with cortical atrophy, reduced brain

12 weight and severity of AD (Montine et al., 1999). Several proteins are involved in maintaining iron homeostatis: these include TfR (iron uptake), Tf (iron transport), Frt (iron storage), HFE (iron uptake), ceruloplasmin (iron transport; feroxidase- conversion of Fe2+ to Fe3+), DMT1 (iron export), and ferroportin (iron export). See Figure 1.1 for schematic showing most key players in iron regulation. Indeed, alterations in the expression pattern of Tf, Frt, and HFE have been reported in the AD brain and will be discussed below. Moreover, genetic mutations in Tf (C2 variant; rs1049296) and two variants in HFE, C282Y (rs1800562) and H63D (rs1799945), shown to affect body iron status (Bartzokis et al.; Fleming et al., 2000; Kauwe et al., 2010), and indices of oxidative stress (Lleo et al., 2002; Loeffler et al., 1995)(), have also been associated with increased risk of AD. There is substantial evidence for the role of iron overload in AD that is further supported by epidemiological evidence implicating variants of iron management proteins in AD risk. The AlzGene meta-analysis of the Tf C2 allele (Bertram et al., 2007) (www.alzgene.org/) currently shows a significant, although low, odds ratio of AD: 1.2 (95% confidence interval, 1.06–1.3) (1 June 2010), with a similar pattern in Caucasians and east Asians. Recently it was also identified in a large genome wide study as a loci significantly associated with AD risk. Evidence of the relation of HFE variants with AD will be discussed below.

Contribution of iron to neurodegeneration in AD has also been proposed to occur by its ability to influence four pathways implicated in AD: 1) APP metabolism (Bodovitz et al., 1995; Kuperstein and Yavin, 2003; Mantyh et al., 1993; Rogers et al., 2002; Rottkamp et al., 2001b; Schubert and Chevion, 1995), 2), the loss of calcium homeostasis (Hidalgo and Nunez, 2007), 3) the degradation of a subset of microglia (Lopes et al., 2008), and 4) oxidative stress (Castellani et al., 2007; Castellani et al., 2012; Honda et al., 2004; Liu et al., 2006; Smith, 2006).

Iron, HFE, and AD

Iron accumulation in the brain is accompanied by an increase in oxidative stress and has been consistently observed in AD (Connor et al., 1992a; Liu et al., 2005; Markesbery, 1997; Maynard et al., 2005). It is believed that iron mediates its toxic effects due to the generation of reactive oxygen species (ROS) by the Fenton reaction (Liu et al., 2006;

13 Okada, 1996). The hemochromatosis (HFE) gene product primarily regulates iron uptake into cells by interacting with the transferrin receptor (TfR) to restrict Tf binding (Feder et al., 1998a). Genetic variants of the HFE gene are unable to maintain iron homeostasis, and in particular, the H63D variant has been under investigation as a potential risk factor for neurodegenerative diseases. Several studies have reported an increased frequency of the H63D mutation in AD patients (Combarros et al., 2003; Connor and Lee, 2006; Moalem et al., 2000; Pulliam et al., 2003; S. Moalem et al., 2000; Sampietro et al., 2001). Pulliam et al. found increased levels of markers of oxidative stress in individual with HFE mutations as compared to controls (Pulliam et al., 2003). It has also been reported that the H63D mutation is five times more frequent in AD patients younger than 70 years compared to patients over 80 (Sampietro et al., 2001). Indeed, MRI studies have revealed increased accumulation of iron in the carriers of HFE mutants (Berg et al., 2000; Nielsen et al., 1995). Several studies have shown that iron staining is most dense in the proximity of amyloid beta (Aβ) plaques and in cells associated with plaques (Connor et al., 1992b) and that iron promotes deposition of Aβ (Rottkamp et al., 2001a). Interestingly, a similar pattern was observed with respect to HFE expression in AD brains, suggesting a normal response of HFE is to limit iron uptake (Connor and Lee, 2006).

HFE gene

HFE (High Iron), originally called HLA-H, is a major histocompatibility complex (MHC) class I-like gene. It is located on the short arm of chromosome 6 and was first identified by Simon and colleagues in association with an iron overload disorder called hereditary hemochromatosis (HH)(Simon et al., 1976). In 1996 Fedar et al identified a mutation (C282Y) in this gene using positional cloning that was present in the homozygous state in nearly 80% of HH cases (Feder et al., 1996a). HH is characterized by tissue cirrhosis and damage to liver, heart, and pancreas resulting from excessive and chronic iron absorption and inability to store iron in iron storage proteins, thereby allowing free iron to cause oxidative damage (Andrews, 1999; Drakesmith and Townsend, 2000).

14 The three most common single nucleotide polymorphisms (SNPs) in the HFE gene are H63D (rs1799945 in exon 2), C282Y (rs1800562 in exon 4), and S65C (Le Gac et al., 2001; Merryweather-Clarke et al., 2000). The H63D variant is caused by a C-to-G transition at nucleotide 187 that results in a histidine to aspartic acid substitution at the 63rd amino acid (Feder et al., 1996a), while the C282Y variant is caused by a G-to-A transition at nucleotide 845 that changes cysteine to tyrosine at the 282nd amino acid (Feder et al., 1996a). The S65C missense mutation results from an A-to T change at nucleotide 193 causing a serine to cysteine substitution at the 65th amino acid. Examination of frequency and global distribution of HFE mutations indicated that HFE variants are most common in Caucasian populations with a frequency of up to 25% for H63D, 12% for C282Y, and 1.6 to 5.5% for S65C. S65C is not as common as other two variants (Camaschella et al., 2002) and has not been associated with physiological changes or diseases (Le Gac et al., 2001).

Not only are the H63D and C282Y variants prevalent, epidemiological evidence associates them with several neurodegenerative, metabolic diseases, and malignancies. Though C282Y is less prevalent in the general population than H63D, it is strongly associated with HH and recently has been under investigation as a risk factor for prostate, breast, colorectal, and brain tumors (Gannon et al., 2011; Osborne et al., 2010; Shaheen et al., 2003; Syrjakoski et al., 2006). It has also been found to be protective for neurodegenerative diseases (Buchanan et al., 2002; Correia et al., 2009). In contrast, the H63D variant has been found to be a risk factor for several diseases such as AD (Combarros et al., 2003; Moalem et al., 2000; Percy et al., 2008a; Pulliam et al., 2003; Sampietro et al., 2001), amyotrophic lateral sclerosis (Goodall et al., 2005; Wang et al., 2004), and stroke (Ellervik et al., 2007b).

The human HFE gene transcript is approximately 4 kilobases (kb) and contains a 1029 nucleotide open reading frame (ORF) encoding a putative polypeptide of 343 residues (Feder et al., 1996a). The mouse gene transcript is approximately 1.5 kb with a 1080 nucleotide ORF that encodes a protein of 359 residues (Hashimoto et al., 1997). The mouse homologue of human HFE is called MR2, and has been found to have significant homology (66%) and exhibit a similar tissue expression pattern as human HFE. The

15 difference in structure between the two arise from the extra eight amino acid residues between the α1 and α2 domains in mouse HFE, that were likely derived from the intron sequence because the DNA sequences in exons encoding the α2 domain align between human and mice (Hashimoto et al., 1997; Riegert et al., 1998). Nonetheless, the evidence of significant iron overload and disruption of iron homeostasis in HFE knockout mice similar to what has been observed in individuals with HH indicates the function of protein is preserved between mice and human and that manipulation of the mouse HFE gene can provide insight into human diseases related to HFE.

Structure of the HFE protein

The HFE gene encodes a 49kDa membrane glycoprotein that resembles MHC-class1 family of proteins. Crystollography studies by Lebron and collegues indicated that the 434 residue single polypeptide HFE protein consists of a transmembrane domain, three extracellular domains (α1 α2 and α3) and a cytoplasmic tail (Bennett et al., 2000; Lebron et al., 1998). The peptide-binding α1‐α2 domains form a platform consisting of eight anti‐ parallel β sheets and two anti-parallel α helices positioned on top of an immunoglobulin- like domain α3 that can bind beta‐2‐microglobulin (β2m) (Lebron et al., 1998). The feature of the HFE protein that resembles MHC class-I proteins is provided by the presence of 4 cysteine residues needed to form disulphide bridges in the α1‐α2 domains needed for the tertiary structure of the protein that allows noncovalent binding to β2m (Bjorkman and Parham, 1990; Feder et al., 1996b). This structural feature is necessary for transport of the HFE protein from the endoplasmic reticulum to the plasma membrane (Miyazaki et al., 1986; Waheed et al., 1997). However, unlike MHC-class I molecules, the surface area of the groove between the two peptide binding domains is much narrower in HFE, limiting its binding to peptides for antigen-presentation (Feder et al., 1997; Lebron et al., 1998). This observation has caused investigators to speculate that HFE may not play a role in immunity even though Rohrlich and colleagues showed that αβ cytolytic T cells without antigen‐presenting function recognized human HFE indicating that it may still function in specific immune responses (Rohrlich et al., 2005). The presence of a cluster of four histidine residues at the base of the α1 domain (region that is mediator of pH-dependent interactions based on its pKa) is a feature of HFE

16 protein structure (Lebron et al., 1998) that if investigated further might result in identification of additional binding partners of the protein and insights into its role in signaling.

The amino acid change in the α1 or α3 domains that result in the H63D and C282Y variants, respectively, significantly affect the structure of HFE. In the H63D variant, substitution of an aspartate at position 63 can disrupt a salt-bridge that is formed between a histidine at the 63rd position and aspartate at the 95th position in WT-HFE, leading to a conformational change (Bennett et al., 2000; Lebron et al., 1998). The replacement of cysteine with a tyrosine in C282Y-HFE disrupts formation of the disulfide bridge necessary for proper folding of the protein needed for its binding to β2M and transport to the cell surface. As a result of this change in conformation, C282Y-HFE cannot leave the ER and move to the plasma membrane. The change in conformation of H63D-HFE still allows its interaction with β2M and cell surface expression but its interaction with other proteins such as TfR is altered resulting in defective function (Feder et al., 1998a).

Function of HFE

The best characterized function of HFE is its role in regulation of cellular iron uptake which it mediates by binding to the transferrin receptor (TfR). Binding affinity of HFE for TfR is comparable to the binding affinity of iron-loaded transferrin (holotransferrin, (Feder et al., 1998b). It has been shown that when a cell has enough iron, HFE competitively binds to the TfR at the site where iron-loaded Tf molecules bind thereby preventing iron uptake via endocytosis. Formation of the HFE-TfR complex is pH- dependent, with a neutral pH allowing tight binding and an acidic pH inhibiting it (Lebron et al., 1998; Lebron et al., 1999). HFE was shown to prevent endocytosis of TfR alone by inducing its phosphorylation. This results in higher cell surface expression of TfR than intracellular expression (Salter-Cid et al., 2000).

The function of HFE in iron regulation is disrupted by mutations causing formation of H63D- and C282Y-HFE variants. It has been shown that while WT-HFE bound with TfR allows only one iron bound Tf (Fe-Tf) to bind per TfR and be taken up by cells (Lebron et al., 1998; Lebron et al., 1999), H63D-HFE allows uptake of more than one molecule of

17 Fe-Tf because it does not reduce affinity of Fe-Tf for TfR. This effect of H63D-HFE is attributed to its slightly abnormal conformation due to the location of the mutation in the α1 domain preventing its efficient binding to TfR (Feder et al., 1997; Waheed et al., 2002). Because C282Y-HFE is retained in the trans-golgi complex due to its inability to bind β2M for cell surface expression, it does not interact with TfR, resulting in more iron uptake than the WT-HFE (Feder et al., 1997; Waheed et al., 2002; Waheed et al., 1999). Overall, both variants are associated with cell and tissue iron-overload. Role of HFE and its variants in iron regulation is depicted below in Figure 1.1 (adapted from Connor and Lee, 2006).

Figure 1.1: Role of HFE and its variants in iron regulation

18 HFE expression

HFE is expressed in a wide range of tissues including liver, intestine, and brain. Parkkila et al. showed that HFE is expressed in gastric epithelial cells, tissue macrophages, and circulating monocytes and granulocytes (Parkkila et al., 2000). Earlier studies (Feder et al., 1996a) did not show expression in brain but (Connor et al., 2001a) showed by immunohistochemistry (IHC) that HFE is expressed in the choroid plexus, endothelial cells of the blood vessels, and ependymal cells lining the ventricles of the brain along with TfR. The expression profile in the brain is consistent with the role of HFE in mediating iron uptake. The association of HFE with TfR and the expression of TfR at the BBB support the idea that iron from serum can be taken up in the brain. In addition, HFE was found to be expressed in neurons surrounding Aβ plaques and NFTs in AD brains (Connor and Lee, 2006; Connor et al., 2001b). The finding of higher expression of HFE in brains of AD patients relative to those of controls indicated that its expression might be induced under conditions of stress (Connor et al., 2001b). Support for this idea was provided by the finding that treatment of BV2 microglial cells with agents that caused an inflammatory response induced expression of HFE (Lee and Connor, 2005).

H63D-HFE and AD

Given that HFE expression was elevated in the brains of individuals with AD and that H63D-HFE resulted in iron overload and elevated oxidative stress, a number of investigators attempted to assess whether H63D-HFE is a risk factor for AD. The factors analyzed to assess the association of H63D-HFE with AD were allelic frequencies and age at disease onset, by gender. A few recent studies also evaluated MCI-to-AD conversion, analysis of above-mentioned factors by age ranges rather than just mean age, and disease risk by interaction between HFE variants and other risk factor alleles such as ApoE4 and the Tf C2 allele. Out of 15 epidemiological studies conducted independently in different populations between 2000 and 2011, 7 studies implicated H63D-HFE as a risk factor in AD based on allelic frequencies, age at onset, or interaction with other genes (Combarros et al., 2003; Correia et al., 2009; Moalem et al., 2000; Percy et al., 2008a; Pulliam et al., 2003; Robson et al., 2004; Sampietro et al., 2001). Notably,

19 Pulliam et al. found increased levels of markers of oxidative stress in individuals with HFE mutations as compared to controls (Pulliam et al., 2003). It has also been reported that the H63D mutation is five times more frequent in AD patients younger than 70 years compared to patients over 80 (Sampietro et al., 2001), suggesting earlier age at disease onset. Three out of 15 studies showed no association of H63D-HFE with AD (Avila- Gomez et al., 2008; Blazquez et al., 2007; Candore et al., 2003). Four additional studies found no association based on allelic frequency but found trends toward a number of factors associated with AD risk. For example, Lleo et al found a trend toward higher frequency of H63D variants in AD (Lleo et al., 2002). Berlin et al. reported a trend of accelerated rate of conversion from MCI to AD in H63D homozygotes (Berlin et al., 2004). Although Guerreiro et al, did not find a significant association for increased risk of AD by allele frequency or age at onset, they did observe an increasing trend for early age of onset in individuals homozygous for H63D (Guerreiro et al., 2006). Interestingly, Alizadeh et al. also observed a trend for early age at onset in H63D homozygous individuals but this effect was more prominent in males with male carriers of both H63D and the ApoE4 allele having a 5.5 year earlier age of disease onset (Alizadeh et al., 2009). The interaction between H63D-HFE and ApoE4 and increased risk of AD in carriers of both mutations was first reported by Moalem et al. and has now been replicated in 4 studies (Alizadeh et al., 2009; Combarros et al., 2003; Moalem et al., 2000; Percy et al., 2008a). Recently, a study of epistatic interactions for AD risk indirectly showed that WT-HFE interacted significantly with another variant of the TF gene (-2AA) to delay onset of AD in Northern Europeans (Lehmann et al., 2012). Another important finding from this study was that carriers of H63D alone had a 2 year earlier disease onset than WT-HFE patients. To account for the smaller sample size in these studies, two meta-analyses were performed. Again, where Ellervik et al. showed a weak association of H63D (1.1 fold increased risk) with AD in 66,000 cases and 226,000 controls (Ellervik et al., 2007a), Lin et al. found a protective effect of H63D in a meta- analysis including 2,795 cases and 7,424 controls (Lin et al., 2012). It is important to note that both studies determined association with AD based on allele frequency in cases and controls and did not consider age at onset, a factor that was consistently found to be significantly altered in AD. In summary, studies showing either positive or negative

20 association have been reported regarding H63D and the risk of AD. However, given the extent of genetic and environmental heterogeneity, contradicting results in epidemiological studies are not considered an anomaly (Gorroochurn et al., 2007). It has been suggested that epistasis interactions could also be the underlying phenomena. Evaluation of cellular and functional consequences of the implicated gene variants in mouse models in the absence of environmental confounding variables, followed by systematic analysis of effects due to other genetic variables and/ or environmental factors in combination with a less penetrant but high susceptibility gene variants like HFE should expand our knowledge.

Nonetheless, cellular expression of H63D-HFE has been shown to result in increased iron accumulation (Lee et al., 2007; Mitchell et al., 2009b), disruption of mitochondrial potential (Lee et al., 2007), increased influx of intracellular Ca2+ (Mitchell et al., 2009b), increased glutamate uptake (Mitchell et al., 2009b), increased secretion of MCP1 that has a role in neuroinflammation (Mitchell et al., 2009a), increased ER stress (Liu et al., 2011), increased oxidative stress (Lee et al., 2007), increased toxicity to Aβ (Mairuae et al., 2010), and decreased Pin1 activity (Hall et al., 2010) contributing to increased Tau phosphorylation (Hall et al., 2011). These findings demonstrate that the expression of H63D-HFE creates a permissive milieu for processes that can influence other pathways in neuronal cells such as lipid homeostasis, neurotransmission, and myelination which can ultimately lead to AD.

22 studies investigated association of C282Y-HFE for risk for AD and none found an association based only single SNP comparisons, rather a protective effect of C282Y-HFE with suggested for AD. In contrast, Robson et al. showed increased risk associated with AD only if individuals carried Tf C2 allele as well as C282Y-HFE (Kauwe et al., 2010; Lehmann et al., 2012; Robson et al., 2004) and this association was further enhanced in tri-carriers: C282Y-HFE, Tf C2, and ApoE4.

21 7. Lipid dyshomeostasis

Lipid rafts

Lipid rafts, also called detergent-resistent microdomains (DRMs), are specialized cell membrane microdomains enriched in cholesterol, sphingolipids, and glycosylphosphatidylinositol (GPI)- anchored proteins. They are thought to affect a number of functions including membrane fluidity, protein sorting, trafficking of membrane proteins during endo- and exocytosis, cell migration, and cell adhesion (Brown and London, 1998; Simons and Toomre, 2000). A cell’s membrane contains both raft and non-raft domains and proteins can interact with either domain. While cholesterol is part of both domains, its enrichment in the raft domains allows it to ‘glue’ other lipids in rafts and create a platform where protein-protein and lipid-protein interactions can occur (Barenholz, 2002). For example, rafts can modulate enzymatic activity by allowing an enzyme and its substrate to come in close contact. Due to this property, lipid rafts can modulate a number of signaling cascades. A number of key receptors and signaling proteins have been found to localize in rafts, while others could be recruited to rafts to regulate cellular processes. It is still not understood how rafts are formed but several models have been proposed. It is generally believed that the inherent affinity of lipids for other lipids and for some proteins drive the lipid-lipid and lipid-protein interactions that result in formation of dynamic ordered lipid domains (Lai, 2003; Lingwood and Simons, 2010). Despite the conflicting opinions about how lipid rafts are formed, it is well- accepted that cholesterol is integral for lipid raft formation and maintenance. Changes in the composition of rafts by depletion of cholesterol or sphingolipids have been shown to alter structure (assembly and disassembly) and function of lipid rafts and ultimately, cellular phenotype (Keller et al., 2004; Veatch and Keller, 2005). Evidence is also accumulating to support the concept that there is a dynamic equilibrium between cholesterol and sphingolipids within the rafts and that when levels of one of these components is altered it affects the concentration of the other within the rafts.

Given the abundance of cholesterol and sphingolipids in the brain, lipid rafts have been proposed as critical for regulating processes such as neuronal signaling, neuronal cell

22 adhesion, axon guidance and neurotransmission required for normal brain function. Lipid rafts are found in neurons, astrocytes, and oligodendrocytes (DeBruin and Harauz, 2007; Gielen et al., 2006; Tsui-Pierchala et al., 2002). Rafts have been shown to facilitate intramembrane proteolysis to activate or inactivate several transmembrane proteins such as APP and receptors involved in neurotophin signaling (Landman and Kim, 2004; Vetrivel et al., 2005). In addition, ionotropic receptors, pre-synaptic SNARE complex proteins (i.e. synaptophysin), as well as, post synaptic receptors (such as NMDA and GABA), and key proteins involved in myelination (i.e. CNPase, MBP) are known to localize in lipid rafts and modulation of lipid raft composition has been shown to affect the function of these proteins (Besshoh et al., 2007; Huang et al., 2007; Marta et al., 2004; Stetzkowski-Marden et al., 2006; Willmann et al., 2006).

The role(s) of lipid rafts in the brain has been under intense investigation, particularly in the context of neurodegenerative diseases. Relevant to AD are studies showing loss of lipid rafts in the temporal cortex (Molander-Melin et al., 2005) and reduced cholesterol in the rafts from hippocampi of AD patients (Abad-Rodriguez et al., 2004; Ledesma, 2003). These studies are consistent with findings of significant total brain cholesterol loss in advanced stage AD patients (Abad-Rodriguez et al., 2004). In addition, a large number of studies have been focused on investigating the role of lipid rafts and cholesterol content of lipid rafts in amyloidogenesis and results have been conflicting. It has been shown that lipid rafts are the sites where APP and presenilin‐1 (enzyme that cleaves APP to generate toxic Aβ) localize or are recruited to promote amyloidogenesis and that a high cholesterol content in rafts favors this process (Bouillot et al., 1996; Hayashi et al., 2000; Lee et al., 1998; Morishima-Kawashima and Ihara, 1998). Therefore, a popular idea has been to try to prevent APP localization to rafts by reducing membrane cholesterol, however, consequences of such an approach on normal APP function are not known. Given that cholesterol/lipid rafts are essential for several other processes, it is possible that this approach may cause more harm than good. Contrary to these observations, it has also been demonstrated in brains of healthy individuals that APP predominantly localizes in non-raft domains and a very small proportion of presenilin-1 is found in rafts. However, in the brains of AD patients APP colocalized with presenilin-1 in non-raft domains.

23 Furthermore, depletion of cholesterol from neuronal cells expressing human APP resulted in colocalization of presenilin-1 APP in non-raft domains and increased Aβ production (Abad-Rodriguez et al., 2004). These findings along with the evidence of low total brain cholesterol in AD patients support the hypothesis that therapeutic approaches to lower brain cholesterol could be detrimental to normal brain function. In fact, accumulating evidence indicates that loss of cholesterol can promote neurodegeneration as lower serum/ total brain/ membrane cholesterol have been implicated in increased risk of Parkinson’s disease (Du et al., 2012), ALS (Albert, 2008; Chio et al., 2009), Smith– Lemli–Opitz syndrome (Jira et al., 2003), Huntington’s (Valenza et al., 2005), Alzheimer’s (Ledesma, 2003; Wellington, 2004; Wolozin, 2004), and Niemann–Pick Type C (Vance, 2006) diseases. NPC disease, which is characterized by reduction of membrane cholesterol and accumulation of cholesterol in the late endosomes and lysosomes of neuronal and other cells (Karten et al., 2009), highlights the importance of membrane cholesterol in normal cells. In summary, lipid rafts modulate a number of cellular processes and neuronal function and alterations in their composition (cholesterol and sphingolipid content) can influence cellular phenotype. The following sections describe sphingolipid and cholesterol metabolism under normal conditions and the contribution of their disruption to AD.

i) Sphingolipids Sphingolipids are a class of lipids that contain sphingosine,which is an aliphatic amino alcohol most often 18 carbons in length with the amino group on C2 and hydroxyls at C1 and C3. Sphingosines are the basic building blocks of sphingolipids which when substituted with additional moieties can be subdivided into simple (ceramides – sphingosine with a fatty acid in an amide linkage on C2) and complex (cerebrosides - containing additional substituents) sphnigolipids. Other types of glycosphingolipids include sphingomyelins, sulfatides, and gangliosides). Sphingolipids are synthesized initially in the ER, sugars added primarily in the Golgi, and found primarily in the plasma membrane in lipid rafts and in endosomes (Futerman, 2006). In addition to cholesterol and sphingomyelin, glycosphingolipids are enriched in rafts. Sphingolipids function in cell-cell recognition, signaling cascades that result in proliferation, apoptosis, stress

24 responses, inflammation, differentiation, and axon growth (Hannun and Obeid, 2002; Lavieu et al., 2006; Snider et al., 2010; Spiegel and Milstien, 2002; Venable et al., 1995).. Unequivocal evidence for the need for appropriate synthesis of sialylated glycosphingolipids was provided by the finding that the brains of children who lacked the ability to synthesize GM3 failed to develop normally (Simpson et al., 2004). Dysregulation of sphingolipid metabolism has also been implicated in a number of metabolic and neurological diseases such as Fabry disease, Krabbe disease, Gaucher disease, Tay-Sachs disease, Metachromatic leukodystrophy, Niemann-Pick disease, AD, and cancer (Haughey et al., 2010; He et al., 2010; Ryland et al., 2011).

Role of sphingolipids in AD

Several studies have shown abnormalities in the lipid content and expression of enzymes regulating their content in AD brain. For example, the total phospholipid and sulfatide content in AD was decreased as compared to normal (Cheng et al., 2003; Gottfries et al., 1996; Han et al., 2002; Pettegrew et al., 2001; Skinner et al., 1989; Soderberg et al., 1992). Sulfatide, produced by oligodendrocytes in the brain and essential for myelination is consistently observed to be deficient in AD (Han et al., 2002). It is not understood what factors cause its depletion. On the other hand, ceramide, a pro-apoptotic lipid has been found to be elevated in the brain (Cutler et al., 2004b) and CSF of patients with AD (Satoi et al., 2005). Ceramide can be produced by hydrolysis of sphingomyelin via sphingomyelinases, or by de novo synthesis from fatty acyl CoA and sphingosine. It has been proposed that oxidative stress and other age-related factors can contribute to age- related accumulation of ceramide and induction of apoptotic signaling in neurons and other cells (Costantini et al., 2005; Cutler et al., 2004b; Kolesnick and Kronke, 1998; Perez et al., 2005). Recently, a study by He et al. replicated the findings of a reduction in sphingomyelin and an elevation of ceramide in AD brains (He et al., 2010). In addition, they also found reduced levels of sphingosine-1-phosphate (S1P), a pro-survival metabolite, which correlated significantly with the levels of Aβ peptide and hyperphosphorylated tau protein (He et al., 2010).

25 In addition to the sphingolipids mentioned above, the relationship of gangliosides, a sub- class of glycosphingolipids, to AD pathogenesis has been investigated. Gangliosides are found in their highest concentration in the gray matter of the brain, with GM1, GD1a, GD1b, and GT1b accounting for 65–85% of that content (Schengrund, 2010). Interestingly, total ganglioside content was found to be reduced in most regions of brains from early-onset or familial cases of AD, while in cases of late-onset or sporadic AD, ganglioside reduction was observed only in the temporal cortex, hippocampus and frontal white matter (Svennerholm and Gottfries, 1994). These observations suggest an age- dependent and/or region-specific pattern of distribution that could be differentially altered in AD.

Though reduction of total brain ganglioside levels in AD has been reported, significant elevation of certain ganglioside species has also been consistently reported. For example, GM1, one of the major ganglioside in brain that is also enriched in lipid rafts, was found to be significantly elevated in brains of AD patients (Svennerholm and Gottfries, 1994) and also in lipid rafts isolated from their frontal and temporal cortices (Molander-Melin et al., 2005).

Additional support for the role of gangliosides in AD was provided by the observation that GM1 stimulated production of Aβ by increasing γ-secretase activity (Zha et al., 2004), that it binds to Aβ with high affinity (Ariga et al., 2001) and could serve as a ‘seed’ for Aβ aggregation by promoting formation of toxic Aβ fibrils (ChooSmith et al., 1997; Kimura and Yanagisawa, 2007; Okada et al., 2007; Yamamoto et al., 2007; Yanagisawa, 2005). Indeed, GM1 was found bound to Aβ in amyloid plaques in AD brains (Hayashi et al., 2004). GM1 has also been shown to influence AD pathogenesis by altering calcium homeostasis. It has been shown that plasma membrane-associated GM1 can enhance intracellular Ca2+ levels by increasing influx of extracellular Ca2+ (Ledeen and Wu, 2002; Wu et al., 2007) which might affect calcium-mediated phosphorylation of tau and APP (Schengrund, 2010). In addition to affecting APP and Ca2+ homeostasis,

26 GM1 has been found to be associated with altered content of other lipids such as cholesterol. Molander-Melin et al. reported an elevation of GM1 and GM2 in lipid rafts isolated from AD brains was associated with a concomitant decrease in their cholesterol (Molander-Melin et al., 2005). Similarly, pharmacological depletion of cholesterol in murine neuroblastoma cells resulted in significant elevation of GM1 (Petro and Schengrund, 2007), suggesting an inverse relationship between GM1 and cholesterol. Collectively, these studies show that sphingolipids play important roles in AD pathogenesis. Further elucidation of factors that can influence sphingolipid composition is needed. Factors inducing oxidative stress such iron are potential targets. Therefore, the effects of H63D-HFE that is associated with iron-overload on sphingolipid metabolism were investigated as described in chapter 3.

ii) Cholesterol metabolism and function in circulation Cholesterol is an essential component of all cell membranes and is needed for maintaining and structure and fluidity of membranes. It is required for growth and replication of mammalian cells, exo- and endocytosis, and is a precursor of steroid hormones and bile acids. Outside the brain, cholesterol is synthesized in the liver through a series of reactions, depicted in Figure 1.2. (Dietschy and Turley, 2001; Dietschy and Turley, 2004). The rate-limiting enzyme in cholesterol biosynthesis is 3-hydroxy-3- methylglutarylcoenzyme A reductase (HMGCoAR) which catalyzes conversion of HMG- CoA to mevalonic acid.

Cholesterol is transported to other organs in the form of lipoproteins with the assistance of transport proteins called apolipoproteins. Based on lipid and protein content, there are five types of lipoproteins found in blood: chylomicrons, very-low-density lipoproteins (VLDL), intermediate-density lipoproteins (IDL), low-density lipoproteins (LDL), and high-density lipoproteins (HDL). Chylomicrons, VLDL, IDL, and LDL deliver cholesterol to cells via the circulation, while HDL remove extra cholesterol from cells and bring it back to the liver where it can be converted into bile acids by the enzymes expressed in liver and excreted from the body. Cells of most organs can either synthesize their own cholesterol or meet their needs by uptake of cholesterol from the circulation.

27 Excess cellular cholesterol is stored in the form of cholesteryl esters in lipid droplets. The esters are formed by the action of an ER-resident enzyme, acyl-coenzyme A cholesterol acyltransferase (ACAT). Cholesterol can also be obtained from dietary sources. Therefore, an efficient system is needed to maintain body cholesterol homeostasis. Indeed, accumulation of cholesterol in arteries leads to blockage of these arteries, restriction of blood flow, and build-up of plaques that ultimately results in a fatal condition called atherosclerosis (Dietschy and Turley, 2001; Lutjohann, 2006). Total plasma cholesterol levels steadily increase with age (20-65 yrs) in both men and women, but have been shown to decrease in men after age 65 (Ettinger et al., 1992) and plateau in women after menopause after initially rising (Brown et al., 1993; Tremollieres et al., 1999). Findings regarding cholesterol content in the plasma in ageing and AD individuals are inconsistent with regards to whether elevated or lower serum cholesterol is a risk factor for AD. A consistent observation in human studies is that high serum cholesterol in mid-life can increase the risk of developing AD later in life, but is associated with improved cognition in elderly individuals (Schreurs, 2010). Similarly, lower total serum cholesterol (Kim et al., 2002) and lower HDL cholesterol were found to be associated with poor cognition in elderly (Atzmon et al., 2002). However, status of brain cholesterol should be a better predictor of AD pathogensis.

Figure 1.2: Cholesterol synthesis pathway

Cholesterol is synthesized by a series of reactions as shown. Iron is a required co-factor for HMGCoAR and CY46A1 (Modified from Bjorkhem 2004).

29 Cholesterol metabolism in the brain

In the brain, almost all of the cholesterol is synthesized in situ, with little to none taken up from the circulation. The insulation of brain from changes in circulating cholesterol is achieved by the presence of the blood brain barrier (BBB). The highest rate of cholesterol synthesis in the CNS occurs during early stages of development especially when myelination is occurring (Jurevics and Morell, 1995; Jurevics and Morell, 1997). The mature brain continues to synthesize cholesterol but at very low level (Boyles et al., 1985; Mcmillan et al., 1972). The finding that the half-life of cholesterol in the adult human brain is very high, approximately 5 years, indicates that an efficient cholesterol recycling mechanism must exist within the brain (Bjorkhem et al., 1998).

In the brain, oligodendrocytes, astrocytes and neurons are capable of synthesizing cholesterol (Saito et al., 1987; Suzuki et al., 2007). Most of the cholesterol synthesized by oligodendrocytes is found in myelin. Neurons also have a very high demand for cholesterol due to membrane turnover in synaptic transmission, however, but it is believed that neurons obtain most of their cholesterol from astrocytes (Srivastava et al., 1997; Stone et al., 1997). So the prevailing view is that in the CNS, astrocytes synthesize cholesterol and with the help of ApoE, also synthesized by astrocytes (Dolev and Michaelson, 2004; LaDu, 1998; Pitas et al., 1987b; Pitas et al., 1987c; Schmechel, 1993), cholesterol is delivered to neurons where it is taken up by cell surface low density lipoprotein receptors (LDLR) via endocytosis. Though other apolipoproteins are also found in CNS i.e Apo AI, ApoJ, ApoD, evidence suggests that ApoE is highly abundant in the brain and is responsible for the shuttling of cholesterol between astrocytes and neurons (Bjorkhem and Meaney, 2004). A well-accepted model to describe cholesterol transport in the brain is shown in Figure 1.3 which is adapted from (Bjorkhem and Meaney, 2004).

In addition to recycling of cholesterol, there is need for removal of excess cholesterol from the brain. Due to its hydrophobic nature, cholesterol cannot cross the BBB. Two

30 mechanisms have been proposed for removal of brain cholesterol; cholesterol hydroxylation and apoE-dependent efflux. The presence of ApoE-bound cholesterol in the CSF indicates that some excretion of brain cholesterol occurs this way but the mechanism is not understood. Based on the rate of CSF renewal and the amount of cholesterol found in the CSF, it has been estimated that 1 to 2 mg cholesterol may be eliminated from the brain via the CSF each day (Pitas et al., 1987a). The major mechanism of cholesterol elimination from the brain is carried out by the cholesterol 24- hydroxylase (CYP46A1)-catalyzed conversion of cholesterol into 24S- hydroxycholesterol (24S-HC), a metabolite that can traverse the BBB (Bjorkhem et al., 1997; Bjorkhem et al., 1998; Lutjohann et al., 1996). This metabolic reaction is particularly important in increasing the ability of 24S-HC (oxysterol) to cross the lipophilic membranes: introduction of the hydroxyl group in the side chain of the cholesterol molecule causes a rearrangement of membrane phospholipids that allows excretion of the oxysterol in an energetically favorable manner (Kessel et al., 2001). This is outlined schematically in Figure 1.4. The direction of movement is mediated by the concentration gradient (Bjorkhem and Meaney, 2004). Because almost all of the 24S-HC originates from the brain, it has been suggested that it can be used as a marker of brain cholesterol homeostasis. Consistent with this view, a number of studies have shown altered levels of 24S-HC in CSF or plasma of patients with neurological diseases such as AD and multiple sclerosis (Bretillon et al., 2000; Leoni et al., 2002; Lutjohann et al., 2000). These studies also highlight the importance of steady and normal flux of cholesterol across the CNS. It has been reported that in healthy humans the flux of cholesterol across the CNS is much less and equals approximately 0.09 mg/day per kg (Saito et al., 1987) and 1.4mg/day per kg in mice (Panzenboeck et al., 2002). However, in humans with AD cholesterol efflux across the CNS is elevated and is proportional to the severity of dementia (Pfrieger, 2003a). A similar trend was observed in a mouse model of Niemen Pick C1 disease (NPC1), where the increase in cholesterol efflux from the brain was associated with increased neurodegeneration (Panzenboeck et al., 2002).

Figure 1.3: Mechanism of cholesterol transport between astrocytes and neurons in the brain. Cholesterol is synthesized by astrocytes and is delivered to neurons. Excess cholesterol is converted into 24S-HC is also an indicator to inhibit cholesterol synthesis in CNS. This model is proposed by Prieger and colleagues and is considered to be a well- accepted mechanism of brain cholesterol transport.

Figure 1.4: Schematic to show redistribution of membrane lipids to allow cholesterol efflux via 24S-HC.

33 Brain vs. plasma cholesterol

With regard to the exchangeability between serum and brain cholesterol, the idea that under normal conditions cholesterol from serum does not cross into the brain due to the BBB, is well-established. Support for this concept first came in the early 1940s from studies showing no incorporation of label into lipids of the brain or spinal cord after administration of D2O to adult rats (Waelsch et al., 1940). These findings have been replicated by several investigators (Dietschy and Turley, 2001; Jurevics and Morell, 1995). Furthermore, studies with intravenous injections of 14C-labeled cholesterol to healthy volunteers and to pregnant women showed no label in the brains of the healthy volunteers (Meaney et al., 2001) or fetal brain tissue (Plotz et al., 1968), emphasizing that almost all of the cholesterol in the brain comes from de novo synthesis with little to none from plasma under normal physiological conditions. Surprisingly, brain endothelial cells have been shown to take up a small amount of LDL cholesterol via LDL receptors expressed on their luminal surface (Dehouck et al., 1994; Dehouck et al., 1997). However, evidence from isotope experiments did not show significant uptake of cholesterol. It has been hypothesized that uptake of small amounts of plasma lipoproteins into the CNS could occur at low levels not detected by current methods (Dietschy and Turley, 2001). Alternatively, significant uptake of cholesterol from plasma could occur under pathological conditions where the integrity of the BBB is compromised or in cases of chronic hypercholesterolemia. This possibility is supported by studies showing a number of metabolic changes in the brain of individuals/experimental animals fed diets high in cholesterol. Currently, knowledge is limited in this respect and further studies are needed to elucidate the effects of dietary cholesterol on brain cholesterol, especially in cases of neurological diseases.

Role of cholesterol in the brain

In the central nervous system, cholesterol is needed for myelination and synaptic transmission. In addition, it is an essential constituent of cellular membranes and lipid raft microdomains, where it can influence signal transduction and cellular processes. The importance of cholesterol in the brain is evident from the fact that brain has ~25% of the bodies cholesterol (Bjorkhem and Meaney, 2004). In the brain, cholesterol is found

34 primarily in myelin (i.e. oligodendroglia) and it is also present in the membranes of all other cell types. Notably, myelin consists of ~70% lipids (mostly cholesterol, phospholipids, and glycosphingolipids i.e. galactocerebroside in molar ratios of ≈4:4:2) and 30% proteins (Bjorkhem and Meaney, 2004). The high lipid content of myelin is consistent with the membrane properties required for its role in supporting saltatory conduction. Myelin is produced by oligodendrocytes (Miller, 2002) and cholesterol enrichment in myelin reduces permeability to ions allowing the action potential to move down the axon without diffusing across the membrane (Bjorkhem and Meaney, 2004). It has been shown that a high cholesterol content is needed for proper myelination (Saher et al., 2005). Disruption of lipid homeostasis significantly alters CNS structure and function. Deficiency in cholesterol synthesis by oligodendrocytes has been shown to delay proper myelination (Marcus and Popko, 2002; Saher et al., 2005). It has also been shown that cholesterol-rich lipid rafts are crucial for the development and maintenance of myelin membranes (Colognato et al., 2002; DeBruin and Harauz, 2007; Gielen et al., 2006).

Several studies have shown that cholesterol is needed for endo- and exocytosis and thereby plays a crucial role in synapse structure and function (Pfrieger, 2003b; Takamori et al., 2006). Consistent with this, studies have shown that cholesterol is enriched in presynaptic terminals and its pharmacological depletion reduced synaptic transmission (Suzuki et al., 2004). Several key synaptic proteins such as synaptophysin and the SNARE proteins are either predominantly found in lipid rafts or must be recruited into rafts in order to effectively interact with other proteins to promote neurotransmitter release and transmission. These effects are thought to be mediated by direct interaction of cholesterol with synaptophysin (Thiele et al., 2000), cholesterol-dependent formation of SNARE complex (Lang et al., 2001; Mitter et al., 2003). Moreover, Suzuki et al. showed that an increase in the cholesterol content of lipid rafts, induced by treatment with brain derived neurotropic factor (BDNF), resulted in increased expression of raft-associated presynaptic proteins, changes associated with synapse development (Suzuki et al., 2007). Collectively, these studies highlight the importance of cholesterol in regulating expression of synaptic proteins and for synaptic transmission.

35 Cholesterol homeostasis and AD pathology

There is substantial epidemiological, genetic, and molecular evidence of disruption of cholesterol homeostasis in AD pathogenesis (Shobab et al., 2005; Sjogren et al., 2006). Mutations in several genes involved in cholesterol uptake, such as LRP and the very-low- density lipoprotein (VLDL) receptor (Zerbinatti and Bu, 2005), as well as in enzymes that regulate cholesterol catabolism such as Cyp46A1 (Vaya and Schipper, 2007; Wolozin, 2003), have been associated with increased risk of AD. In addition, epidemiological studies revealed increased susceptibility to AD in patients with elevated plasma cholesterol levels in mid-life (Jarvik et al., 1995; Notkola et al., 1998; Roher et al., 1999). However, in comparison to healthy brains, cholesterol levels are much lower in the brains of patients who had AD when they died (Ledesma, 2003; Ledesma and Dotti, 2006). This finding not only indicates that the regulation of brain cholesterol is distinct from plasma; it also suggests that in AD dysregulation of brain cholesterol can result in a greater loss of cholesterol. Excessive loss of brain cholesterol can be particularly devastating because the rate of brain cholesterol synthesis decreases with age and under normal conditions there is no influx of cholesterol from the plasma into the brain. Therefore, the brain must be able to recycle cholesterol and lipoproteins such as APOE in order to prevent undo loss of cholesterol from the brain that can lead to neurodegeneration.

Cholesterol in aging brain

Changes in brain structure and function have been consistently observed with increasing age (Teter and Finch, 2004) and ageing is the biggest risk factor for AD. Unlike plasma cholesterol, assessment of brain cholesterol in humans is challenging due to our inability to detect it in the brains of living individuals. Studies of cholesterol in brains obtained from animals with AD symptoms and human brain autopsy tissue have yielded inconclusive results regarding whether high or low cholesterol is a risk factor for AD. The lack of agreement between studies could be due to analysis of cholesterol content from different regions/domains i.e. total brain cholesterol vs. membrane cholesterol vs. cholesterol in lipid raft domains. It has been reported that while the asymmetric distribution of cholesterol in the plasma membrane of aged mice was reduced

36 relative to younger mice, total membrane cholesterol was unaltered (Igbavboa et al., 1996). However, cholesterol content of lipid rafts from brains of older mice expressing either human ApoE3 or ApoE4 was greater than in younger animals (Igbavboa et al., 2005). It has also been reported that brain cholesterol increases in certain conditions with age i.e. NPC1 (Vance, 2006) while other studies provided evidence for decline in total brain cholesterol i.e. AD (Ledesma, 2003)and Huntington’s disease (Valenza et al., 2005). In addition to total brain cholesterol, maintenance of cellular cholesterol content is crucial in mediating APP processing via alpha and gamma secretases (Bogdanovic et al., 2002; Cam and Bu, 2006) and therefore it has physiological relevance to AD. Ginsberg et al. have shown that in AD brains cell membranes are less stable due to altered lipid composition (Ginsberg et al., 1993a; Ginsberg et al., 1993b). Nonetheless, mechanistic experimental studies have shown that either abnormally high or low cholesterol can be associated with pathogenic manifestations seen in neurological diseases.

Cholesterol content of the membrane regulates its fluidity with greater amounts of cholesterol making them more rigid and lower cholesterol making them more fluid and allowing more permeability to ions (Barenholz, 2002). Therefore, disruption of cholesterol distribution in the membrane can affect its function in maintaining normal cell behavior. Consistent with this view, altered dendritic morphology was observed upon depletion of membrane cholesterol in cultured neurons (Holtzman et al., 1995). Because red blood cells (RBCs) come in close contact with certain regions of the brain, it is thought that the changes in the membranes of RBC could reflect alterations in the brain and such alterations could be used as biomarkers to diagnose and monitor progression of neurological diseases. In fact, morphology of RBCs and the protein composition of RBC membranes from RBCs isolated from AD subjects have been shown to be altered (Mohanty et al., 2010; Sabolovic et al., 1997). Furthermore, oxidative stress associated changes in the phospholipid content were observed in RBC membranes isolated from RBCs from AD patients (Oma et al., 2012).

Taken together, the above-mentioned studies indicate that alterations in cholesterol homeostasis can affect multiple processes that may affect neurodegeneration,

37 thereby influencing cognitive ability. Modulation of cholesterol levels thus becomes a potential target for the prevention or treatment of neurodegenerative conditions. In this regard, knowledge of genetic and environmental factors that can influence cholesterol metabolism should aid in the development of better therapeutics.

ApoE functions and isoform-specific effects relevant to AD pathology

ApoE is a major apolipoprotein and a cholesterol carrier in brain (Mahley, 1988). In the brain, apoE is synthesized and secreted primarily from astrocytes as part of small dense cholesterol containing lipoproteins (Dolev and Michaelson, 2004; LaDu, 1998; Pitas et al., 1987b; Pitas et al., 1987c; Schmechel, 1993). It is responsible for carrying cholesterol to neurons for regenerative functions (Srivastava et al., 1997; Stone et al., 1997). ApoE binds to the low density lipoprotein receptor (LDLR) and LDLR-related protein (LRP1) and taken up by neurons via receptor-mediated endocytosis (Herz and Bock, 2002; Herz and Chen, 2006). Because there is no influx of ApoE and cholesterol from the plasma to brain due to the blood brain barrier, both are synthesized locally in the brain (Baskin et al., 1997; Boschert et al., 1999; Cedazo-Minguez et al., 2001; Ignatius et al., 1986) and must be retained in the CNS to maintain cholesterol homeostasis. Therefore, maintenance of efficient cholesterol recycling via ApoE is thought to be crucial for cholesterol homeostasis in the brain. It is not clear what factors may cause disruption of brain cholesterol homeostasis seen in association with AD pathology.

In humans, ApoE exists as three isoforms (E2, E3, E4) (Mahley, 1988). The association of ApoE4 as a strong risk factor for AD has been well established (Corder, 1993; Mahley, 1988; Mahley et al., 2006). Despite advances in knowledge of how ApoE functions,, the molecular mechanisms whereby ApoE4 dysfunction leads to AD are not completely understood. In the general population, ApoE3 is the most common isoform (allele frequency 77-78%), followed by ApoE4 (15%) (Mahley, 1988). However, ApoE4 is present in ~40% of AD patients (Corder, 1993). The three isoforms E2, E3 and E4 differ from one another at residues 112 and 158 and these differences result in differential effects on lipid and receptor binding properties and other cellular processes (Mahley,

38 1988; Mahley et al., 2006). Consistent with increased risk of AD in carriers of the ApoE4 allele, expression of ApoE4 protein in vitro and in vivo has been shown to be more deleterious than expression of the ApoE3 isoform because ApoE4 is shown to be more prone to aggregation and is unable to perform a number of functions efficiently including cholesterol transport, Aβ clearance, and antioxidation (Mahley et al., 2006).

The role of APOE in AD pathology has been heavily investigated in regards to its ability to influence Aβ metabolism. Several lines of evidence indicate that CNS-derived APOE-containing lipoprotein particles may not directly influence Aβ synthesis but promote its clearance and degradation by binding to it (DeMattos, 2004; Koistinaho et al., 2004; Manelli et al., 2004; Stratman et al., 2005; Trommer et al., 2005). In addition, increased Aβ burden has been consistently reported in ApoE4 carriers, in both sporadic (Schmechel, 1993) and genetic AD cases (Bogdanovic et al., 2002), as well as in mouse models (Dolev and Michaelson, 2004; Hartman et al., 2002). Recently, in vivo and in vitro studies by Riddell et al. showed that astrocytes specifically degraded ApoE4, causing a reduction in ApoE secretion and brain levels (Riddell, 2008). Collectively, these studies emphasize the protective role of ApoE3 in Aβ clearance and suggest a loss of function for ApoE4.

ApoE isoforms have also been reported to differentially influence tau pathology with ApoE3 found to prevent its phosphorylation by binding the non-phosphorylated form in vitro (Strittmater, 1994). In contrast, increased phosphorylation of tau was observed in mice expressing human apoE4 (Brecht, 2004; Tesseur, 2000). Interestingly, accumulation of hyperphosphorylated tau was also found in ApoE knock-out mice fed a high cholesterol diet (Rahman et al., 2005), indicating the need for normal ApoE to maintain cholesterol homeostasis and appropriate tau phosphorylation.

Under physiological conditions, neurons do not express ApoE, but depend on ApoE supplied by astrocytes. Neurons take up astrocyte-released ApoE via LRP1 mediated endocytosis (Herz, 1988; Jeon and Blacklow, 2005). Importance of the role of APOE in distributing cholesterol within the cell membrane was made evident by the observation of Igbav-boa et al., showing that apoE-deficient mice had twice as much cholesterol in the extracellular side of the synaptic plasma membrane than the

39 intracellular layer (Igbavboa et al., 1997). In addition, isoform-specific effects in cholesterol transport were observed with ApoE4 being less efficient than ApoE3 (Gong, 2002; Michikawa et al., 2000; Rapp et al., 2006). There is also evidence that ApoE3- expressing astrocytes can supply more cholesterol to neurons than ApoE4 expressing astrocytes (Gong, 2002). Recent reports about ability of neurons to recycle ApoE and retain ApoE inside cells (Fazio et al., 2000) have led to the hypothesis that ApoE may play additional roles in mediating cell signaling and intracellular homeostasis. One such role is its function as an antioxidant which was found to be due to its ability to bind certain metals such as copper, iron, and zinc, with the highest binding affinity being for iron (Miyata and Smith, 1996) . Interestingly, the antioxidant and metal binding activities of ApoE were found to be allele-specific, with ApoE4 less efficient in binding metals (Mutter et al., 2004) and reducing oxidative stress than ApoE3 (Miyata and Smith, 1996).

Rationale for thesis

Despite the efforts that have been made to identify factors that can trigger the pathological events associated with AD we still don’t have a clear understanding of what they are. It has long been believed that accumulation of Aβ is the triggering event, preceding NFT formation, and ultimately culminating in synaptic failure and memory impairment. However, evidence is also available indicating that tau pathology may precede formation of amyloid plaques. Indeed, evidence of memory decline with symptoms characteristic of AD, without amyloid plaques and/ or NFTs, leads to questioning whether these pathological markers are the causative agents. Recently, evidence has started to accumulate in support of the idea that formation of Aβ may be a protective event designed to counteract additional insults in the cells. Since molecular phenotypes such as oxidative stress, synaptic failure, neuronal loss, and cognitive decline, characteristics associated with AD, have been shown to result from disruption of a number of the pathways discussed above, one can easily argue that there is no linear sequence of events that results in AD. Therefore, a multi targeted approach is needed to study a complex disorder like AD and to develop effective therapeutics. This can be achieved only when knowledge about interactions between the different pathways and the potential influence of environmental factors on them becomes available. A good example

40 of an interconnection between pathways that may have relevance to AD pathogenesis lies at the intersection between iron and cholesterol metabolism. Both iron and cholesterol metabolism have been independently implicated in the etiology of AD as described above. A large number of studies have established that disruption of cholesterol metabolism is associated with multiple aspects of AD, including APP metabolism, Tau phosphorylation, synaptic integrity and transmission, and cognitive function. Iron dyshomeostasis has also been shown to contribute to the above-mentioned aspects of AD, suggesting a possible crosstalk between the two pathways that may result in synergistic deleterious effects contributing to the development of AD. How these pathways may influence each other is currently unknown. Evidence is accumulating for a potential link between iron and cholesterol metabolism in the context of atherosclerosis, yet this link and its relevance to AD remains largely unexplored. There is some evidence to support this association. 1) Iron is a required cofactor for a number of enzymes involved in cholesterol metabolism so it is logical to speculate that alteration in iron content could influence the activity of those enzymes. 2) Iron, ApoE, and cholesterol have all been found in association with extracellular amyloid plaques and intracellular neurofibrillary tangles, hallmark features of AD. 3) Iron is a potent source of oxidative radicals which have been consistently reported to affect membrane lipids and thereby influence lipid homeostasis. 4) Rabbits fed a cholesterol-rich diet were found to accumulate iron and Aβ deposits in the brain, and had increased mortality (Ghribi et al., 2006). Finally, and perhaps most relevant are the recent findings that carriers of both the ApoE4 (cholesterol transporter) allele and H63D-HFE variant (iron accumulation) had increased risk and earlier onset of AD (Combarros et al., 2003; Moalem et al., 2000; Percy et al., 2008b; Sampietro et al., 2001). Collectively, these studies highlight an area of investigation that could lead to a better understanding of the underlying causes for AD.

Therefore, the overall goal of this thesis was to interrogate the link between iron and cholesterol and its relevance to diseases. In particular, studies presented in this thesis focused on understanding how disruption of iron homeostasis could influence cholesterol and sphingolipid metabolism to affect cellular, molecular, and behavioral processes characteristic of AD. The association of iron and cholesterol dyshomeostasis to cancer is

41 also addressed in this thesis with background information for that topic provided in chapter 4.

The global hypothesis under investigation in my thesis is that expression of HFE variants establishes a permissive cellular milieu for pathogenic processes. There are two areas of investigation in this thesis that focus on the two most prevalent HFE gene variants. For the studies on neurological function, we have focused on the H63D gene variant because this is the allelic variant seen most frequently in late onset neurodegenerative diseases. The specific hypothesis for these studies is that expression of H63D-HFE is associated with altered cholesterol and sphingolipid metabolism which can lead to disruption of lipid raft-mediated signaling pathways, culminating in memory impairment characteristic of AD. The second area of investigation in my thesis focuses on the C282Y allelic variant and cancer. The hypothesis for these studies is that the expression of C282Y-HFE is associated with disruption of cholesterol and sphingolipid metabolism. To address the first hypothesis, I used novel cell culture and mouse models of iron overload due to expression of H63D-HFE as well as ones that expressed WT-HFE as controls and had normal iron content. In chapter 2, the effects of H63D-HFE on cholesterol metabolism and phenotypes associated with AD were investigated. Chapter 3 addressed the impact of HFE variants (H63D and C282Y) on sphingolipid metabolism. The second hypothesis was investigated using a novel cell culture model stably expressing C282Y-HFE and compare with cells expressing WT-HFE. The effects of C282Y-HFE on cholesterol metabolism and implication of the resulting alterations on cancer are explored in chapter 4. Lastly, because of the importance of cholesterol in membrane integrity and cellular function, studies were conducted to investigate the lipid composition of red blood cell (erythrocyte) membranes in children with autism spectrum disorder. Findings from this study are described in the chapter that is part of appendix.

42 References:

Abad-Rodriguez, J., M.D. Ledesma, K. Craessaerts, S. Perga, M. Medina, A. Delacourte, C. Dingwall, B. De Strooper, and C.G. Dotti. 2004. Neuronal membrane cholesterol loss enhances amyloid peptide generation. J Cell Biol. 167:953-60.

Abraham, R., V. Moskvina, R. Sims, P. Hollingworth, A. Morgan, L. Georgieva, K. Dowzell, S. Cichon, A.M. Hillmer, M.C. O'Donovan, J. Williams, M.J. Owen, and G. Kirov. 2008. A genome-wide association study for late-onset Alzheimer's disease using DNA pooling. BMC Med Genomics. 1:44.

Adalbert, R., J. Gilley, and M.P. Coleman. 2007. A[beta], tau and ApoE4 in Alzheimer's disease: the axonal connection. Trends in Molecular Medicine. 13:135-142.

Albert, S.M. 2008. Dyslipidemia in ALS - Good, bad, or unclear? Neurology. 70:988- 989.

Alessenko, A.V., A.E. Bugrova, and L.B. Dudnik. 2004. Connection of lipid peroxide oxidation with the sphingomyelin pathway in the development of Alzheimer's disease. Biochemical Society Transactions. 32:144-146.

Alizadeh, B.Z., O.T. Njajou, M.R. Millan, A. Hofman, M.M. Breteler, and C.M. van Duijn. 2009. HFE variants, APOE and Alzheimer's disease: findings from the population-based Rotterdam study. Neurobiol Aging. 30:330-2.

Alzheimer, A., R.A. Stelzmann, H.N. Schnitzlein, and F.R. Murtagh. 1995. An English translation of Alzheimer's 1907 paper, "Uber eine eigenartige Erkankung der Hirnrinde". Clin Anat. 8:429-31.

Amadoro, G., V. Corsetti, A. Atlante, F. Florenzano, S. Capsoni, R. Bussani, D. Mercanti, and P. Calissano. 2012. Interaction between NH2-tau fragment and A beta in Alzheimer's disease mitochondria contributes to the synaptic deterioration. Neurobiology of Aging. 33.

Andrews, N.C. 1999. Disorders of iron metabolism. N Engl J Med. 341:1986-95.

43 Arendt, T. 2009. Synaptic degeneration in Alzheimer's disease. Acta Neuropathologica. 118:167-179.

Ariga, T., K. Kobayashi, A. Hasegawa, M. Kiso, H. Ishida, and T. Miyatake. 2001. Characterization of high-affinity binding between gangliosides and amyloid beta- protein. Arch Biochem Biophys. 388:225-30.

Atzmon, G., I. Gabriely, W. Greiner, D. Davidson, C. Schechter, and N. Barzilai. 2002. Plasma HDL levels highly correlate with cognitive function in exceptional longevity. J Gerontol A Biol Sci Med Sci. 57:M712-5.

Avila-Gomez, I.C., M. Jimenez-Del-Rio, F. Lopera-Restrepo, and C. Velez-Pardo. 2008. Association between HFE 187 C > G (H63D) mutation and early-onset familial Alzheimer's disease PSEN-1 839A > C (E280A) mutation. Annals of Hematology. 87:671-673.

Ballard, C., S. Gauthier, A. Corbett, C. Brayne, D. Aarsland, and E. Jones. 2011. Alzheimer's disease. Lancet. 377:1019-31.

Bao, F.X., L. Wicklund, P.N. Lacor, W.L. Klein, A. Nordberg, and A. Marutle. 2012. Different beta-amyloid oligomer assemblies in Alzheimer brains correlate with age of disease onset and impaired cholinergic activity. Neurobiology of Aging. 33.

Barenholz, Y. 2002. Cholesterol and other membrane active sterols: from membrane evolution to "rafts". Progress in Lipid Research. 41:1-5.

Barsoum, M.J., H. Yuan, A.A. Gerencser, G. Liot, Y. Kushnareva, S. Graber, I. Kovacs, W.D. Lee, J. Waggoner, J. Cui, A.D. White, B. Bossy, J.C. Martinou, R.J. Youle, S.A. Lipton, M.H. Ellisman, G.A. Perkins, and E. Bossy-Wetzel. 2006. Nitric oxide-induced mitochondrial fission is regulated by dynamin-related GTPases in neurons. EMBO J. 25:3900-11.

Bartzokis, G. 2009. Alzheimer's disease as homeostatic responses to age-related myelin breakdown. Neurobiol Aging. 32:1341-71.

44 Bartzokis, G., P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P. Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely, and J.R. Connor. Prevalent iron metabolism gene variants associated with increased brain ferritin iron in healthy older men. J Alzheimers Dis. 20:333-41.

Bartzokis, G., J. Mintz, D. Sultzer, P. Marx, J.S. Herzberg, C.K. Phelan, and S.R. Marder. 1994. In vivo MR evaluation of age-related increases in brain iron. AJNR Am J Neuroradiol. 15:1129-38.

Bartzokis, G., D. Sultzer, J. Cummings, L.E. Holt, D.B. Hance, V.W. Henderson, and J. Mintz. 2000. In vivo evaluation of brain iron in Alzheimer disease using magnetic resonance imaging. Arch Gen Psychiatry. 57:47-53.

Bartzokis, G., T.A. Tishler, P.H. Lu, P. Villablanca, L.L. Altshuler, M. Carter, D. Huang, N. Edwards, and J. Mintz. 2007. Brain ferritin iron may influence age- and gender-related risks of neurodegeneration. Neurobiol Aging. 28:414-23.

Baskin, F., G.M. Smith, J.A. Fosmire, and R.N. Rosenberg. 1997. Altered apolipoprotein E secretion in cytokine treated human astrocyte cultures. Journal of the Neurological Sciences. 148:15-18.

Bennett, M.J., J.A. Lebron, and P.J. Bjorkman. 2000. Crystal structure of the hereditary hemochromatosis protein HFE complexed with transferrin receptor. Faseb Journal. 14:A482-A482.

Berchtold, N.C., and C.W. Cotman. 1998. Evolution in the conceptualization of dementia and Alzheimer's disease: Greco-Roman period to the 1960s. Neurobiol Aging. 19:173-89.

Berg, D., U. Hoggenmuller, E. Hofmann, R. Fischer, M. Kraus, M. Scheurlen, and G. Becker. 2000. The basal ganglia in haemochromatosis. Neuroradiology. 42:9-13.

Berlin, D., G. Chong, H. Chertkow, H. Bergman, N.A. Phillips, and H.M. Schipper. 2004. Evaluation of HFE (hemochromatosis) mutations as genetic modifiers in sporadic AD and MCI. Neurobiol Aging. 25:465-74.

45 Bertram, L., C. Lange, K. Mullin, M. Parkinson, M. Hsiao, M.F. Hogan, B.M. Schjeide, B. Hooli, J. Divito, I. Ionita, H. Jiang, N. Laird, T. Moscarillo, K.L. Ohlsen, K. Elliott, X. Wang, D. Hu-Lince, M. Ryder, A. Murphy, S.L. Wagner, D. Blacker, K.D. Becker, and R.E. Tanzi. 2008. Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE. Am J Hum Genet. 83:623-32.

Bertram, L., M.B. McQueen, K. Mullin, D. Blacker, and R.E. Tanzi. 2007. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nature Genetics. 39:17-23.

Besshoh, S., S. Chen, I.R. Brown, and J.W. Gurd. 2007. Developmental changes in the association of NMDA receptors with lipid rafts. Journal of Neuroscience Research. 85:1876-1883.

Bjorkhem, I., D. Lutjohann, O. Breuer, A. Sakinis, and A. Wennmalm. 1997. Importance of a novel oxidative mechanism for elimination of brain cholesterol - Turnover of cholesterol and 24(S)-hydroxycholesterol in rat brain as measured with O-18(2) techniques in vivo and in vitro. Journal of Biological Chemistry. 272:30178- 30184.

Bjorkhem, I., D. Lutjohann, U. Diczfalusy, L. Stahle, G. Ahlborg, and J. Wahren. 1998. Cholesterol homeostasis in human brain: turnover of 24S-hydroxycholesterol and evidence for a cerebral origin of most of this oxysterol in the circulation. J Lipid Res. 39:1594-600.

Bjorkhem, I., and S. Meaney. 2004. Brain cholesterol: Long secret life behind a barrier. Arteriosclerosis Thrombosis and Vascular Biology. 24:806-815.

Bjorkman, P.J., and P. Parham. 1990. Structure, Function, and Diversity of Class-I Major Histocompatibility Complex-Molecules. Annual Review of Biochemistry. 59:253- 288.

Blazquez, L., D. De Juan, J. Ruiz-Martinez, J.I. Emparanza, A. Saenz, D. Otaegui, A. Sistiaga, P. Martinez-Lage, I. Lamet, L. Samaranch, C. Buiza, I. Etxeberria, E.

46 Arriola, E. Cuadrado, E. Urdaneta, J. Yanguas, and A. Lopez de Munain. 2007. Genes related to iron metabolism and susceptibility to Alzheimer's disease in Basque population. Neurobiol Aging. 28:1941-3.

Bodovitz, S., M.T. Falduto, D.E. Frail, and W.L. Klein. 1995. Iron Levels Modulate Alpha-Secretase Cleavage of Amyloid Precursor Protein. Journal of Neurochemistry. 64:307-315.

Bogdanovic, N., E. Corder, L. Lannfelt, and B. Winblad. 2002. APOE polymorphism and clinical duration determine regional neuropathology in Swedish APP(670, 671) mutation carriers: implications for late-onset Alzheimer's disease. J. Cell. Mol. Med. 6:199-214.

Boschert, U., E. Merlo-Pich, G. Higgins, A.D. Roses, and S. Catsicas. 1999. Apolipoprotein E expression by neurons surviving excitotoxic stress. Neurobiol Dis. 6:508-14.

Bouillot, C., A. Prochiantz, G. Rougon, and B. Allinquant. 1996. Axonal amyloid precursor protein expressed by neurons in vitro is present in a membrane fraction with caveolae-like properties. Journal of Biological Chemistry. 271:7640-7644.

Boyles, J.K., R.E. Pitas, E. Wilson, R.W. Mahley, and J.M. Taylor. 1985. Apolipoprotein-E Associated with Astrocytic Glia of the Central Nervous-System and with Nonmyelinating Glia of the Peripheral Nervous-System. Journal of Clinical Investigation. 76:1501-1513.

Braak, H., and E. Braak. 1995. Staging of Alzheimer's disease-related neurofibrillary changes. Neurobiol Aging. 16:271-8; discussion 278-84.

Brecht, W.J. 2004. Neuron-specific apolipoprotein e4 proteolysis is associated with increased tau phosphorylation in brains of transgenic mice. J. Neurosci. 24:2527- 2534.

Bretillon, L., D. Lutjohann, L. Stahle, T. Widhe, L. Bindl, G. Eggertsen, U. Diczfalusy, and I. Bjorkhem. 2000. Plasma levels of 24S-hydroxycholesterol reflect the

47 balance between cerebral production and hepatic metabolism and are inversely related to body surface. J Lipid Res. 41:840-5.

Brown, D.A., and E. London. 1998. Functions of lipid rafts in biological membranes. Annual Review of Cell and Developmental Biology. 14:111-136.

Brown, S.A., R. Hutchinson, J. Morrisett, E. Boerwinkle, C.E. Davis, A.M. Gotto, and W. Patsch. 1993. Plasma-Lipid, Lipoprotein Cholesterol, and Apoprotein Distributions in Selected United-States Communities - the Atherosclerosis Risk in Communities (Aric) Study. Arteriosclerosis and Thrombosis. 13:1139-1158.

Brys, M., E. Pirraglia, K. Rich, S. Rolstad, L. Mosconi, R. Switalski, L. Glodzik- Sobanska, S. De Santi, R. Zinkowski, P. Mehta, D. Pratico, L.A. Saint Louis, A. Wallin, K. Blennow, and M.J. de Leon. 2009. Prediction and longitudinal study of CSF biomarkers in mild cognitive impairment. Neurobiology of Aging. 30:682- 690.

Buchanan, D.D., P.A. Silburn, J.B. Chalk, D.G. Le Couteur, and G.D. Mellick. 2002. The Cys282Tyr polymorphism in the HFE gene in Australian Parkinson's disease patients. Neuroscience Letters. 327:91-94.

Bush, A.I. 2003. The metallobiology of Alzheimer's disease. Trends Neurosci. 26:207-14.

Cai, Z.Y., B. Zhao, and A. Ratka. 2011. Oxidative Stress and beta-Amyloid Protein in Alzheimer's Disease. Neuromolecular Medicine. 13:223-250.

Callaway, E. 2012. Alzheimer's drugs take a new tack. Nature. 489:13-4.

Cam, J.A., and G. Bu. 2006. Modulation of beta-amyloid precursor protein trafficking and processing by the low density lipoprotein receptor family. Mol Neurodegener. 1:8.

Camaschella, C., A. Roetto, and M. De Gobbi. 2002. Genetic haemochromatosis: genes and mutations associated with iron loading. Best Practice & Research Clinical Haematology. 15:261-276.

48 Candore, G., F. Licastro, M. Chiappelli, C. Franceschi, D. Lio, C. Rita Balistreri, G. Piazza, G. Colonna-Romano, L.M. Grimaldi, and C. Caruso. 2003. Association between the HFE mutations and unsuccessful ageing: a study in Alzheimer's disease patients from Northern Italy. Mech Ageing Dev. 124:525-8.

Canevari, L., and J.B. Clark. 2007. Alzheimer's disease and cholesterol: the fat connection. Neurochem Res. 32:739-50.

Carrasquillo, M.M., F. Zou, V.S. Pankratz, S.L. Wilcox, L. Ma, L.P. Walker, S.G. Younkin, C.S. Younkin, L.H. Younkin, G.D. Bisceglio, N. Ertekin-Taner, J.E. Crook, D.W. Dickson, R.C. Petersen, and N.R. Graff-Radford. 2009. Genetic variation in PCDH11X is associated with susceptibility to late-onset Alzheimer's disease. Nat Genet. 41:192-8.

Castellani, R.J., P.I. Moreira, G. Liu, J. Dobson, G. Perry, M.A. Smith, and X.W. Zhu. 2007. Iron: The redox-active center of oxidative stress in Alzheimer disease. Neurochemical Research. 32:1640-1645.

Castellani, R.J., P.I. Moreira, G. Perry, and X.W. Zhu. 2012. The role of iron as a mediator of oxidative stress in Alzheimer disease. Biofactors. 38:133-138.

Cedazo-Minguez, A., U. Hamker, V. Meske, R.W. Veh, R. Hellweg, C. Jacobi, F. Albert, R.F. Cowburn, and T.G. Ohm. 2001. Regulation of apolipoprotein E secretion in rat primary hippocampal astrocyte cultures. Neuroscience. 105:651-61.

Cheng, H., J. Xu, D.W. McKeel, and X. Han. 2003. Specificity and potential mechanism of sulfatide deficiency in Alzheimer's disease: An electrospray ionization mass spectrometric study. Cellular and Molecular Biology. 49:809-818.

Chio, A., A. Calvo, A. Ilardi, E. Cavallo, C. Moglia, R. Mutani, A. Palmo, R. Galletti, K. Marinou, L. Papetti, and G. Mora. 2009. Lower serum lipid levels are related to respiratory impairment in patients with ALS. Neurology. 73:1681-1685.

49 Cho, D.H., T. Nakamura, J. Fang, P. Cieplak, A. Godzik, Z. Gu, and S.A. Lipton. 2009. S-nitrosylation of Drp1 mediates beta-amyloid-related mitochondrial fission and neuronal injury. Science. 324:102-5.

ChooSmith, L.P., W. GarzonRodriguez, C.G. Glabe, and W.K. Surewicz. 1997. Acceleration of amyloid fibril formation by specific binding of A beta-(1-40) peptide to ganglioside-containing membrane vesicles. Journal of Biological Chemistry. 272:22987-22990.

Coleman, P.D., and P.J. Yao. 2003. Synaptic slaughter in Alzheimer's disease. Neurobiology of Aging. 24:1023-1027.

Colognato, H., W. Baron, V. Avellana-Adalid, J.B. Relvas, A. Baron-Van Evercooren, E. Georges-Labouesse, and C. ffrench-Constant. 2002. CNS integrins switch growth factor signalling to promote target-dependent survival. Nature Cell Biology. 4:833-841.

Combarros, O., M. Garcia-Roman, A. Fontalba, J.L. Fernandez-Luna, J. Llorca, J. Infante, and J. Berciano. 2003. Interaction of the H63D mutation in the hemochromatosis gene with the apolipoprotein E epsilon 4 allele modulates age at onset of Alzheimer's disease. Dement Geriatr Cogn Disord. 15:151-4.

Connor, J.R., and S.Y. Lee. 2006. HFE mutations and Alzheimer's disease. J Alzheimers Dis. 10:267-76.

Connor, J.R., S.L. Menzies, J.R. Burdo, and P.J. Boyer. 2001a. Iron and iron management proteins in neurobiology. Pediatr Neurol. 25:118-29.

Connor, J.R., S.L. Menzies, S.M. St Martin, and E.J. Mufson. 1992a. A histochemical study of iron, transferrin, and ferritin in Alzheimer's diseased brains. J Neurosci Res. 31:75-83.

Connor, J.R., E.A. Milward, S. Moalem, M. Sampietro, P. Boyer, M.E. Percy, C. Vergani, R.J. Scott, and M. Chorney. 2001b. Is hemochromatosis a risk factor for Alzheimer's disease? J Alzheimers Dis. 3:471-477.

50 Connor, J.R., B.S. Snyder, P. Arosio, D.A. Loeffler, and P. Lewitt. 1995. A Quantitative- Analysis of Isoferritins in Select Regions of Aged, Parkinsonian, and, Alzheimers Diseased Brains. Journal of Neurochemistry. 65:717-724.

Connor, J.R., B.S. Snyder, J.L. Beard, R.E. Fine, and E.J. Mufson. 1992b. Regional distribution of iron and iron-regulatory proteins in the brain in aging and Alzheimer's disease. J Neurosci Res. 31:327-35.

Coon, K.D., A.J. Myers, D.W. Craig, J.A. Webster, J.V. Pearson, D.H. Lince, V.L. Zismann, T.G. Beach, D. Leung, L. Bryden, R.F. Halperin, L. Marlowe, M. Kaleem, D.G. Walker, R. Ravid, C.B. Heward, J. Rogers, A. Papassotiropoulos, E.M. Reiman, J. Hardy, and D.A. Stephan. 2007. A high-density whole-genome association study reveals that APOE is the major susceptibility gene for sporadic late-onset Alzheimer's disease. J Clin Psychiatry. 68:613-8.

Corder, E.H. 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science. 261:921-923.

Corder, E.H., A.M. Saunders, W.J. Strittmatter, D.E. Schmechel, P.C. Gaskell, G.W. Small, A.D. Roses, J.L. Haines, and M.A. Pericak-Vance. 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science. 261:921-3.

Correia, A.P., J.P. Pinto, V. Dias, C. Mascarenhas, S. Almeida, and G. Porto. 2009. CAT53 and HFE alleles in Alzheimer's disease: A putative protective role of the C282Y HFE mutation. Neuroscience Letters. 457:129-132.

Costantini, C., R. Weindruch, G. Della Valle, and L. Puglielli. 2005. TrkA-to-p75(NTR) molecular switch activates amyloid beta-peptide generation during aging. Biochemical Journal. 391:59-67.

Cutler, R.G., J. Kelly, K. Storie, W.A. Pedersen, A. Tammara, K. Hatanpaa, J.C. Troncoso, and M.P. Mattson. 2004a. Involvement of oxidative stress-induced abnormalities in ceramide and cholesterol metabolism in brain aging and

51 Alzheimer's disease. Proceedings of the National Academy of Sciences of the United States of America. 101:2070-2075.

Cutler, R.G., J. Kelly, K. Storie, W.A. Pedersen, A. Tammara, K. Hatanpaa, J.C. Troncoso, and M.P. Mattson. 2004b. Involvement of oxidative stress-induced abnormalities in ceramide and cholesterol metabolism in brain aging and Alzheimer's disease. Proc Natl Acad Sci U S A. 101:2070-5.

De Jonghe, C., C. Zehr, D. Yager, C.M. Prada, S. Younkin, L. Hendriks, C. Van Broeckhoven, and C.B. Eckman. 1998. Flemish and Dutch mutations in amyloid beta precursor protein have different effects on amyloid beta secretion. Neurobiology of Disease. 5:281-286.

DeBruin, L.S., and G. Harauz. 2007. White matter rafting - Membrane microdomains in myelin. Neurochemical Research. 32:213-228.

Decuypere, J.P., G. Monaco, G. Bultynck, L. Missiaen, H. De Smedt, and J.B. Parys. 2010. The IP(3) receptor-mitochondria connection in apoptosis and autophagy. Biochim Biophys Acta. 1813:1003-13.

Dehouck, B., M.P. Dehouck, J.C. Fruchart, and R. Cecchelli. 1994. Upregulation of the low density lipoprotein receptor at the blood-brain barrier: intercommunications between brain capillary endothelial cells and astrocytes. J Cell Biol. 126:465-73.

Dehouck, B., L. Fenart, M.P. Dehouck, A. Pierce, G. Torpier, and R. Cecchelli. 1997. A new function for the LDL receptor: transcytosis of LDL across the blood-brain barrier. J Cell Biol. 138:877-89.

DeMattos, R.B. 2004. ApoE and clusterin cooperatively suppress A[beta] levels and deposition: evidence that ApoE regulates extracellular A[beta] metabolism in vivo. Neuron. 41:193-202.

Dietschy, J.M., and S.D. Turley. 2001. Cholesterol metabolism in the brain. Curr Opin Lipidol. 12:105-12.

52 Dietschy, J.M., and S.D. Turley. 2004. Thematic review series: brain Lipids. Cholesterol metabolism in the central nervous system during early development and in the mature animal. J Lipid Res. 45:1375-97.

Ding, B., K.M. Chen, H.W. Ling, F. Sun, X. Li, T. Wan, W.M. Chai, H. Zhang, Y. Zhan, and Y.J. Guan. 2009. Correlation of Iron in the Hippocampus with MMSE in Patients with Alzheimer's Disease. Journal of Magnetic Resonance Imaging. 29:793-798.

Dolev, I., and D.M. Michaelson. 2004. A nontransgenic mouse model shows inducible amyloid-beta (Abeta) peptide deposition and elucidates the role of apolipoprotein E in the amyloid cascade. Proc Natl Acad Sci U S A. 101:13909-14.

Drakesmith, H., and A. Townsend. 2000. The structure and function of HFE. Bioessays. 22:595-8.

Du, G.W., M.M. Lewis, M.L. Shaffer, H.L. Chen, Q.X. Yang, R.B. Mailman, and X.M. Huang. 2012. Serum Cholesterol and Nigrostriatal R2*Values in Parkinson's Disease. PLoS One. 7.

Ellervik, C., H. Birgens, A. Tybjaerg-Hansen, and B.G. Nordestgaard. 2007a. Hemochromatosis genotypes and risk of 31 disease endpoints meta-analyses including 66,000 cases and 226,000 controls. Haematologica-the Hematology Journal. 92:59-59.

Ellervik, C., A. Tybjaerg-Hansen, M. Appleyard, H. Sillesen, G. Boysen, and B.G. Nordestgaard. 2007b. Hereditary hemochromatosis genotypes and risk of ischemic stroke. Neurology. 68:1025-1031.

Ermak, G., and K.J.A. Davies. 2002. Calcium and oxidative stress: from cell signaling to cell death. Molecular Immunology. 38:713-721.

Ettinger, W.H., P.W. Wahl, L.H. Kuller, T.L. Bush, R.P. Tracy, T.A. Manolio, N.O. Borhani, N.D. Wong, D.H. Oleary, C.D. Furberg, M.E. Bond, G. Heiss, S. Klopfenstein, M. Lyles, M. Mittelmark, G.S. Tell, J.F. Toole, M. Cody, G.

53 Garner, G. Cruise, J. Robbins, W. Bommer, M. Lee, M.B. Schenker, C.J. Tupper, T. Himmelmann, F. Labaw, J. Kay, P. Borhani, L.P. Fried, G.W. Comstock, P.S. German, S.J. Kittner, S. Kumanyika, T.R. Price, R.C. Rock, R.N. Bryan, M. Szklo, B. Tabatznik, M.S. Tockman, J. Hill, J.B. Chabot, J. Cauley, K. Matthews, A. Newman, T.J. Orchard, G.H. Rutan, R. Schulz, V.E. Smith, S.K. Wolfson, P. Meyer, B. Mclaughlin, J.M. Gardin, H. Antonculver, W.L. Henry, S. Lobodzinski, J. Whittenberger, P. Knoll, J. Conner, J.F. Polak, J. Potter, E. Bovill, E. Cornell, P. Enright, S. Toogood, P. Rautaharju, F. Rautaharju, R.A. Kronmal, B.M. Psaty, D. Siscovick, B. Hermanson, P.J. Savage, and P. Smith. 1992. Lipoprotein Lipids in Older-People - Results from the Cardiovascular Health Study. Circulation. 86:858-869.

Farooqui, A.A., L.A. Horrocks, and T. Farooqui. 2007. Interactions between neural membrane glycerophospholipid and sphingolipid mediators: A recipe for neural cell survival or suicide. Journal of Neuroscience Research. 85:1834-1850.

Fazio, S., M.F. Linton, and L.L. Swift. 2000. The Cell Biology and Physiologic Relevance of ApoE Recycling. Trends in Cardiovascular Medicine. 10:23-30.

Feder, J.N., A. Gnirke, W. Thomas, Z. Tsuchihashi, D.A. Ruddy, A. Basava, F. Dormishian, R. Domingo, M.C. Ellis, A. Fullan, L.M. Hinton, N.L. Jones, B.E. Kimmel, G.S. Kronmal, P. Lauer, V.K. Lee, D.B. Loeb, F.A. Mapa, E. McClelland, N.C. Meyer, G.A. Mintier, N. Moeller, T. Moore, E. Morikang, C.E. Prass, L. Quintana, S.M. Starnes, R.C. Schatzman, K.J. Brunke, D.T. Drayna, N.J. Risch, B.R. Bacon, and R.K. Wolff. 1996a. A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nature Genetics. 13:399- 408.

Feder, J.N., A. Gnirke, W. Thomas, Z. Tsuchihashi, D.A. Ruddy, N.J. Risch, B.R. Bacon, and R.K. Wolff. 1996b. A novel MHC class I-like gene is mutated in patients with hemochromatosis. Blood. 88:2579-2579.

Feder, J.N., D.M. Penny, A. Irrinki, V.K. Lee, J.A. LebrÃ³n, N. Watson, Z. Tsuchihashi, E. Sigal, P.J. Bjorkman, and R.C. Schatzman. 1998a. The hemochromatosis gene

54 product complexes with the transferrin receptor and lowers its affinity for ligand binding. Proceedings of the National Academy of Sciences of the United States of America. 95:1472-1477.

Feder, J.N., D.M. Penny, A. Irrinki, V.K. Lee, J.A. Lebron, N. Watson, Z. Tsuchihashi, E. Sigal, P.J. Bjorkman, and R.C. Schatzman. 1998b. The hemochromatosis gene product complexes with the transferrin receptor and lowers its affinity for ligand binding. Proc Natl Acad Sci U S A. 95:1472-7.

Feder, J.N., Z. Tsuchihashi, A. Irrinki, V.K. Lee, F.A. Mapa, E. Morikang, C.E. Prass, S.M. Starnes, R.K. Wolff, S. Parkkila, W.S. Sly, and R.C. Schatzman. 1997. The hemochromatosis founder mutation in HLA-H disrupts beta(2)-microglobulin interaction and cell surface expression. Journal of Biological Chemistry. 272:14025-14028.

Ferrari, C.K.B. 2000. Free radicals, lipid peroxidation and antioxidants in apoptosis: implications in cancer, cardiovascular and neurological diseases. Biologia. 55:581-590.

Fleming, R.E., M.C. Migas, C.C. Holden, A. Waheed, R.S. Britton, S. Tomatsu, B.R. Bacon, and W.S. Sly. 2000. Transferrin receptor 2: continued expression in mouse liver in the face of iron overload and in hereditary hemochromatosis. Proc Natl Acad Sci U S A. 97:2214-9.

Futerman, A.H. 2006. Intracellular trafficking of sphingolipids: Relationship to biosynthesis. Biochimica Et Biophysica Acta-Biomembranes. 1758:1885-1892.

Gannon, P.O., S. Medelci, C. Le Page, M. Beaulieu, D.M. Provencher, A.M. Mes- Masson, and M.M. Santos. 2011. Impact of hemochromatosis gene (HFE) mutations on epithelial ovarian cancer risk and prognosis. International Journal of Cancer. 128:2326-2334.

Gatz, M., C.A. Reynolds, L. Fratiglioni, B. Johansson, J.A. Mortimer, S. Berg, A. Fiske, and N.L. Pedersen. 2006. Role of genes and environments for explaining Alzheimer disease. Arch Gen Psychiatry. 63:168-74.

55 Ghribi, O., M.Y. Golovko, B. Larsen, M. Schrag, and E.J. Murphy. 2006. Deposition of iron and beta-amyloid plaques is associated with cortical cellular damage in rabbits fed with long-term cholesterol-enriched diets. J Neurochem. 99:438-49.

Gielen, E., W. Baron, M. Vandeven, P. Steels, D. Hoekstra, and M. Ameloot. 2006. Rafts in oligodendrocytes: Evidence and structure-function relationship. Glia. 54:499- 512.

Ginsberg, L., J.R. Atack, S.I. Rapoport, and N.L. Gershfeld. 1993a. Evidence for a membrane lipid defect in Alzheimer disease. Mol Chem Neuropathol. 19:37-46.

Ginsberg, L., J.R. Atack, S.I. Rapoport, and N.L. Gershfeld. 1993b. Regional specificity of membrane instability in Alzheimer's disease brain. Brain Res. 615:355-7.

Glenner, G.G., and C.W. Wong. 1984. Alzheimer's disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun. 120:885-90.

Goate, A., M.C. Chartier-Harlin, M. Mullan, J. Brown, F. Crawford, L. Fidani, L. Giuffra, A. Haynes, N. Irving, L. James, and et al. 1991. Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature. 349:704-6.

Gong, J.S. 2002. Apolipoprotein E (ApoE) isoform-dependent lipid release from astrocytes prepared from human ApoE3 and ApoE4 knock-in mice. J. Biol. Chem. 277:29919-29926.

Goodall, E.F., M.J. Greenway, I. van Marion, C.B. Carroll, O. Hardiman, and K.E. Morrison. 2005. Association of the H63D polymorphism in the hemochromatosis gene with sporadic ALS. Neurology. 65:934-937.

Gorroochurn, P., S.E. Hodge, G.A. Heiman, M. Durner, and D.A. Greenberg. 2007. Non- replication of association studies: "pseudo-failures" to replicate? Genet Med. 9:325-31.

56 Gottfries, C.G., I. Karlsson, and L. Svennerholm. 1996. Membrane components separate early-onset Alzheimer's disease from senile dementia of the Alzheimer type. Int Psychogeriatr. 8:365-72.

Gotz, J., A. Schild, F. Hoerndli, and L. Pennanen. 2004. Amyloid-induced neurofibrillary tangle formation in Alzheimer's disease: insight from transgenic mouse and tissue-culture models. Int J Dev Neurosci. 22:453-65.

Grundman, M., R.C. Petersen, S.H. Ferris, R.G. Thomas, P.S. Aisen, D.A. Bennett, N.L. Foster, C.R. Jack, D.R. Galasko, R. Doody, J. Kaye, M. Sano, R. Mohs, S. Gauthier, H.T. Kim, S. Jin, A.N. Schultz, K. Schafer, R. Mulnard, C.H. van Dyck, J. Mintzer, E.Y. Zamrini, D. Cahn-Weiner, L.J. Thal, and A.s.D.C. Study. 2004. Mild cognitive impairment can be distinguished from Alzheimer disease and normal aging for clinical trials. Archives of Neurology. 61:59-66.

Grupe, A., R. Abraham, Y. Li, C. Rowland, P. Hollingworth, A. Morgan, L. Jehu, R. Segurado, D. Stone, E. Schadt, M. Karnoub, P. Nowotny, K. Tacey, J. Catanese, J. Sninsky, C. Brayne, D. Rubinsztein, M. Gill, B. Lawlor, S. Lovestone, P. Holmans, M. O'Donovan, J.C. Morris, L. Thal, A. Goate, M.J. Owen, and J. Williams. 2007. Evidence for novel susceptibility genes for late-onset Alzheimer's disease from a genome-wide association study of putative functional variants. Hum Mol Genet. 16:865-73.

Guerreiro, R.J., J.M. Bras, I. Santana, C. Januario, B. Santiago, A.S. Morgadinho, M.H. Ribeiro, J. Hardy, A. Singleton, and C. Oliveira. 2006. Association of HFE common mutations with Parkinson's disease, Alzheimer's disease and mild cognitive impairment in a Portuguese cohort. BMC Neurol. 6:24.

Haass, C., C.A. Lemere, A. Capell, M. Citron, P. Seubert, D. Schenk, L. Lannfelt, and D.J. Selkoe. 1995. The Swedish Mutation Causes Early-Onset Alzheimers- Disease by Beta-Secretase Cleavage within the Secretory Pathway. Nature Medicine. 1:1291-1296.

57 Hall, E.C., 2nd, S.Y. Lee, Z. Simmons, E.B. Neely, W. Nandar, and J.R. Connor. 2010. Prolyl-peptidyl isomerase, Pin1, phosphorylation is compromised in association with the expression of the HFE polymorphic allele, H63D. Biochim Biophys Acta. 1802:389-95.

Hall, E.C., S.Y. Lee, N. Mairuae, Z. Simmons, and J.R. Connor. 2011. Expression of the HFE allelic variant H63D in SH-SY5Y cells affects tau phosphorylation at serine residues. Neurobiology of Aging. 32:1409-1419.

Hallgren, B., and P. Sourander. 1958. The Effect of Age on the Non-Haemin Iron in the Human Brain. Journal of Neurochemistry. 3:41-51.

Han, X.L., D.M. Holtzman, D.W. McKeel, J. Kelley, and J.C. Morris. 2002. Substantial sulfatide deficiency and ceramide elevation in very early Alzheimer's disease: potential role in disease pathogenesis. Journal of Neurochemistry. 82:809-818.

Hannun, Y.A., and L.M. Obeid. 2002. The ceramide-centric universe of lipid-mediated cell regulation: Stress encounters of the lipid kind. Journal of Biological Chemistry. 277:25847-25850.

Harold, D., R. Abraham, P. Hollingworth, R. Sims, A. Gerrish, M.L. Hamshere, J.S. Pahwa, V. Moskvina, K. Dowzell, A. Williams, N. Jones, C. Thomas, A. Stretton, A.R. Morgan, S. Lovestone, J. Powell, P. Proitsi, M.K. Lupton, C. Brayne, D.C. Rubinsztein, M. Gill, B. Lawlor, A. Lynch, K. Morgan, K.S. Brown, P.A. Passmore, D. Craig, B. McGuinness, S. Todd, C. Holmes, D. Mann, A.D. Smith, S. Love, P.G. Kehoe, J. Hardy, S. Mead, N. Fox, M. Rossor, J. Collinge, W. Maier, F. Jessen, B. Schurmann, H. van den Bussche, I. Heuser, J. Kornhuber, J. Wiltfang, M. Dichgans, L. Frolich, H. Hampel, M. Hull, D. Rujescu, A.M. Goate, J.S. Kauwe, C. Cruchaga, P. Nowotny, J.C. Morris, K. Mayo, K. Sleegers, K. Bettens, S. Engelborghs, P.P. De Deyn, C. Van Broeckhoven, G. Livingston, N.J. Bass, H. Gurling, A. McQuillin, R. Gwilliam, P. Deloukas, A. Al-Chalabi, C.E. Shaw, M. Tsolaki, A.B. Singleton, R. Guerreiro, T.W. Muhleisen, M.M. Nothen, S. Moebus, K.H. Jockel, N. Klopp, H.E. Wichmann, M.M. Carrasquillo, V.S. Pankratz, S.G. Younkin, P.A. Holmans, M. O'Donovan, M.J. Owen, and J.

58 Williams. 2009. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nat Genet. 41:1088-93.

Hartl, F.U., A. Bracher, and M. Hayer-Hartl. 2011. Molecular chaperones in protein folding and proteostasis. Nature. 475:324-32.

Hartman, R.E., H. Laurer, L. Longhi, K.R. Bales, S.M. Paul, T.K. McIntosh, and D.M. Holtzman. 2002. Apolipoprotein E4 influences amyloid deposition but not cell loss after traumatic brain injury in a mouse model of Alzheimer's disease. J Neurosci. 22:10083-7.

Hartmann, T., J. Kuchenbecker, and M.O.W. Grimm. 2007. Alzheimer's disease: the lipid connection. Journal of Neurochemistry. 103:159-170.

Hashimoto, K., M. Hirai, and Y. Kurosawa. 1997. Identification of a mouse homolog for the human hereditary haemochromatosis candidate gene. Biochemical and Biophysical Research Communications. 230:35-39.

Haughey, N.J., V.V.R. Bandaru, M. Bae, and M.P. Mattson. 2010. Roles for dysfunctional sphingolipid metabolism in Alzheimer's disease neuropathogenesis. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids. 1801:878- 886.

Hayashi, H., N. Kimura, H. Yamaguchi, K. Hasegawa, T. Yokoseki, M. Shibata, N. Yamamoto, M. Michikawa, Y. Yoshikawa, K. Terao, K. Matsuzaki, C.A. Lemere, D.J. Selkoe, H. Naiki, and K. Yanagisawa. 2004. A seed for Alzheimer amyloid in the brain. Journal of Neuroscience. 24:4894-4902.

Hayashi, H., T. Mizuno, M. Michikawa, C. Haass, and K. Yanagisawa. 2000. Amyloid precursor protein in unique cholesterol-rich microdomains different from caveolae-like domains. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids. 1483:81-90.

59 He, X.X., Y. Huang, B. Li, C.X. Gong, and E.H. Schuchman. 2010. Deregulation of sphingolipid metabolism in Alzheimer's disease. Neurobiology of Aging. 31:398- 408.

Heinzen, E.L., A.C. Need, K.M. Hayden, O. Chiba-Falek, A.D. Roses, W.J. Strittmatter, J.R. Burke, C.M. Hulette, K.A. Welsh-Bohmer, and D.B. Goldstein. 2012. Genome-wide scan of copy number variation in late-onset Alzheimer's disease. J Alzheimers Dis. 19:69-77.

Herz, J. 1988. Surface location and high affinity for calcium of a 500-kd liver membrane protein closely related to the LDL-receptor suggest a physiological role as lipoprotein receptor. EMBO J. 7:4119-4127.

Herz, J., and H.H. Bock. 2002. Lipoprotein receptors in the nervous system. Annu. Rev. Biochem. 71:405-434.

Herz, J., and Y. Chen. 2006. Reelin, lipoprotein receptors and synaptic plasticity. Nature Rev. Neurosci. 7:850-859.

Hidalgo, C., and M.T. Nunez. 2007. Calcium, iron and neuronal function. Iubmb Life. 59:280-285.

Hirose, W., K. Ikematsu, and R. Tsuda. 2003. Age-associated increases in heme oxygenase-1 and ferritin immunoreactivity in the autopsied brain. Leg Med (Tokyo). 5 Suppl 1:S360-6.

Hollingworth, P., D. Harold, L. Jones, M.J. Owen, and J. Williams. 2011a. Alzheimer's disease genetics: current knowledge and future challenges. Int J Geriatr Psychiatry. 26:793-802.

Hollingworth, P., D. Harold, R. Sims, A. Gerrish, J.C. Lambert, M.M. Carrasquillo, R. Abraham, M.L. Hamshere, J.S. Pahwa, V. Moskvina, K. Dowzell, N. Jones, A. Stretton, C. Thomas, A. Richards, D. Ivanov, C. Widdowson, J. Chapman, S. Lovestone, J. Powell, P. Proitsi, M.K. Lupton, C. Brayne, D.C. Rubinsztein, M. Gill, B. Lawlor, A. Lynch, K.S. Brown, P.A. Passmore, D. Craig, B. McGuinness,

60 S. Todd, C. Holmes, D. Mann, A.D. Smith, H. Beaumont, D. Warden, G. Wilcock, S. Love, P.G. Kehoe, N.M. Hooper, E.R. Vardy, J. Hardy, S. Mead, N.C. Fox, M. Rossor, J. Collinge, W. Maier, F. Jessen, E. Ruther, B. Schurmann, R. Heun, H. Kolsch, H. van den Bussche, I. Heuser, J. Kornhuber, J. Wiltfang, M. Dichgans, L. Frolich, H. Hampel, J. Gallacher, M. Hull, D. Rujescu, I. Giegling, A.M. Goate, J.S. Kauwe, C. Cruchaga, P. Nowotny, J.C. Morris, K. Mayo, K. Sleegers, K. Bettens, S. Engelborghs, P.P. De Deyn, C. Van Broeckhoven, G. Livingston, N.J. Bass, H. Gurling, A. McQuillin, R. Gwilliam, P. Deloukas, A. Al-Chalabi, C.E. Shaw, M. Tsolaki, A.B. Singleton, R. Guerreiro, T.W. Muhleisen, M.M. Nothen, S. Moebus, K.H. Jockel, N. Klopp, H.E. Wichmann, V.S. Pankratz, S.B. Sando, J.O. Aasly, M. Barcikowska, Z.K. Wszolek, D.W. Dickson, N.R. Graff-Radford, R.C. Petersen, et al. 2011b. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer's disease. Nat Genet. 43:429-35.

Holtzman, D.M., R.E. Pitas, J. Kilbridge, B. Nathan, R.W. Mahley, G. Bu, and A.L. Schwartz. 1995. Low density lipoprotein receptor-related protein mediates apolipoprotein E-dependent neurite outgrowth in a central nervous system-derived neuronal cell line. Proc Natl Acad Sci U S A. 92:9480-4.

Honda, K., G. Casadesus, R.B. Petersen, G. Perry, and M.A. Smith. 2004. Oxidative stress and redox-active iron in Alzheimer's disease. Redox-Active Metals in Neurological Disorders. 1012:179-182.

Huang, P., W. Xu, S.I. Yoon, C.G. Chen, P.L.G. Chong, and L.Y. Liu-Chen. 2007. Cholesterol reduction by methyl-beta-cyclodextrin attenuates the delta opioid receptor-mediated signaling in neuronal cells but enhances it in non-neuronal cells. Biochemical Pharmacology. 73:534-549.

Hyman, B.T., G.W. Vanhoesen, K. Beyreuther, and C.L. Masters. 1989. A4 Amyloid Protein Immunoreactivity Is Present in Alzheimers-Disease Neurofibrillary Tangles. Neuroscience Letters. 101:352-355.

61 Igbavboa, U., N.A. Avdulov, S.V. Chochina, and W.G. Wood. 1997. Transbilayer distribution of cholesterol is modified in brain synaptic plasma membranes of knockout mice deficient in the low-density lipoprotein receptor, apolipoprotein E, or both proteins. J Neurochem. 69:1661-7.

Igbavboa, U., N.A. Avdulov, F. Schroeder, and W.G. Wood. 1996. Increasing age alters transbilayer fluidity and cholesterol asymmetry in synaptic plasma membranes of mice. Journal of Neurochemistry. 66:1717-1725.

Igbavboa, U., G.P. Eckert, T.M. Malo, A.E. Studniski, L.N.A. Johnson, N. Yamamoto, M. Kobayashi, S.C. Fujita, T.R. Appel, W. Muller, W.G. Wood, and K. Yanagisawa. 2005. Murine synaptosomal lipid raft protein and lipid composition are altered by expression of human apoE 3 and 4 and by increasing age. Journal of the Neurological Sciences. 229:225-232.

Ignatius, M.J., P.J. Gebicke-Harter, J.H. Skene, J.W. Schilling, K.H. Weisgraber, R.W. Mahley, and E.M. Shooter. 1986. Expression of apolipoprotein E during nerve degeneration and regeneration. Proc Natl Acad Sci U S A. 83:1125-9.

Ingelsson, M., H. Fukumoto, K.L. Newell, J.H. Growdon, E.T. Hedley-Whyte, M.P. Frosch, M.S. Albert, B.T. Hyman, and M.C. Irizarry. 2004. Early Abeta accumulation and progressive synaptic loss, gliosis, and tangle formation in AD brain. Neurology. 62:925-31.

Jarvik, G.P., E.M. Wijsman, W.A. Kukull, G.D. Schellenberg, C. Yu, and E.B. Larson. 1995. Interactions of apolipoprotein E genotype, total cholesterol level, age, and sex in prediction of Alzheimer's disease: a case-control study. Neurology. 45:1092-6.

Jeon, H., and S.C. Blacklow. 2005. Structure and physiologic function of the low-density lipoprotein receptor. Annu. Rev. Biochem. 74:535-562.

Jira, P.E., H.R. Waterham, R.J.A. Wanders, J.A.M. Smeitink, R.C.A. Sengers, and R.A. Wevers. 2003. Smith-Lemli-Opitz syndrome and the DHCR7 gene. Annals of Human Genetics. 67:269-280.

62 Jurevics, H., and P. Morell. 1995. Cholesterol for Synthesis of Myelin Is Made Locally, Not Imported into Brain. Journal of Neurochemistry. 64:895-901.

Jurevics, H., and P. Morell. 1997. Sources of cholesterol during development of the rat fetus and fetal organs. Journal of Neurochemistry. 69:S15-S15.

Karten, B., K.B. Peake, and J.E. Vance. 2009. Mechanisms and consequences of impaired lipid trafficking in Niemann-Pick type C1-deficient mammalian cells. Biochimica Et Biophysica Acta-Molecular and Cell Biology of Lipids. 1791:659- 670.

Kauwe, J.S.K., S. Bertelsen, K. Mayo, C. Cruchaga, R. Abraham, P. Hollingworth, D. Harold, M.J. Owen, J. Williams, S. Lovestone, J.C. Morris, A.M. Goate, and A.D.N. Initia. 2010. Suggestive Synergy Between Genetic Variants in TF and HFE as Risk Factors for Alzheimer's Disease. American Journal of Medical Genetics Part B-Neuropsychiatric Genetics. 153B:955-959.

Keller, R.K., T.P. Arnold, and S.J. Fliesler. 2004. Formation of 7-dehydrocholesterol- containing membrane rafts in vitro and in vivo, with relevance to the Smith- Lemli-Opitz syndrome. Journal of Lipid Research. 45:347-355.

Kessel, A., N. Ben-Tal, and S. May. 2001. Interactions of cholesterol with lipid bilayers: The preferred configuration and fluctuations. Biophysical Journal. 81:643-658.

Kim, J.M., R. Stewart, I.S. Shin, and J.S. Yoon. 2002. Low cholesterol, cognitive function and Alzheimer s disease in a community population with cognitive impairment. J Nutr Health Aging. 6:320-3.

Kimura, N., and K. Yanagisawa. 2007. Endosomal accumulation of GM1 ganglioside- bound amyloid beta-protein in neurons of aged monkey brains. Neuroreport. 18:1669-73.

Kispal, G., K. Sipos, H. Lange, Z. Fekete, T. Bedekovics, T. Janaky, J. Bassler, D.J.A. Netz, J. Balk, C. Rotte, and R. Lill. 2005. Biogenesis of cytosolic ribosomes

63 requires the essential iron-sulphur protein Rli1p and mitochondria. Embo Journal. 24:589-598.

Klingenberg, M. 2009. Cardiolipin and mitochondrial carriers. Biochim Biophys Acta. 1788:2048-58.

Koistinaho, M., S. Lin, X. Wu, M. Esterman, D. Koger, J. Hanson, R. Higgs, F. Liu, S. Malkani, K.R. Bales, and S.M. Paul. 2004. Apolipoprotein E promotes astrocyte colocalization and degradation of deposited amyloid-[beta] peptides. Nat Med. 10:719-726.

Kolesnick, R.N., and M. Kronke. 1998. Regulation of ceramide production and apoptosis. Annual Review of Physiology. 60:643-665.

Koudinov, A.R., T.T. Berezov, and N.V. Koudinova. 2002. Cholesterol and Alzheimer's disease: is there a link? Neurology. 58:1135.

Koudinov, A.R., and N.V. Koudinova. 2001. Essential role for cholesterol in synaptic plasticity and neuronal degeneration. FASEB J. 15:1858-60.

Koudinov, A.R., and N.V. Koudinova. 2003. Cholesterol, synaptic function and Alzheimer's disease. Pharmacopsychiatry. 36 Suppl 2:S107-12.

Koudinov, A.R., and N.V. Koudinova. 2005. Cholesterol homeostasis failure as a unifying cause of synaptic degeneration. J Neurol Sci. 229-230:233-40.

Kuperstein, F., and E. Yavin. 2003. Pro-apoptotic signaling in neuronal cells following iron and amyloid beta peptide neurotoxicity. Journal of Neurochemistry. 86:114- 125.

Lackner, L.L., and J. Nunnari. 2010. Small molecule inhibitors of mitochondrial division: tools that translate basic biological research into medicine. Chem Biol. 17:578-83.

LaDu, M.J. 1998. Nascent astrocyte particles differ from lipoproteins in CSF. J. Neurochem. 70:2070-2081.

64 LaFerla, F.M. Pathways linking Abeta and tau pathologies. Biochem Soc Trans. 38:993- 5.

LaFerla, F.M., K.N. Green, and S. Oddo. 2007. Intracellular amyloid-beta in Alzheimer's disease. Nat Rev Neurosci. 8:499-509.

LaFerla, F.M., and S. Oddo. 2005. Alzheimer's disease: Abeta, tau and synaptic dysfunction. Trends Mol Med. 11:170-6.

Lai, E.C. 2003. Lipid rafts make for slippery platforms. Journal of Cell Biology. 162:365- 370.

Lambert, J.C., S. Heath, G. Even, D. Campion, K. Sleegers, M. Hiltunen, O. Combarros, D. Zelenika, M.J. Bullido, B. Tavernier, L. Letenneur, K. Bettens, C. Berr, F. Pasquier, N. Fievet, P. Barberger-Gateau, S. Engelborghs, P. De Deyn, I. Mateo, A. Franck, S. Helisalmi, E. Porcellini, O. Hanon, M.M. de Pancorbo, C. Lendon, C. Dufouil, C. Jaillard, T. Leveillard, V. Alvarez, P. Bosco, M. Mancuso, F. Panza, B. Nacmias, P. Bossu, P. Piccardi, G. Annoni, D. Seripa, D. Galimberti, D. Hannequin, F. Licastro, H. Soininen, K. Ritchie, H. Blanche, J.F. Dartigues, C. Tzourio, I. Gut, C. Van Broeckhoven, A. Alperovitch, M. Lathrop, and P. Amouyel. 2009. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nat Genet. 41:1094-9.

Landman, N., and T.W. Kim. 2004. Got RIP? Presenilin-dependent intramembrane proteolysis in growth factor receptor signaling. Cytokine & Growth Factor Reviews. 15:337-351.

Lang, T., D. Bruns, D. Wenzel, D. Riedel, P. Holroyd, C. Thiele, and R. Jahn. 2001. SNAREs are concentrated in cholesterol-dependent clusters that define docking and fusion sites for exocytosis. Embo Journal. 20:2202-2213.

Lange, S.J., and L. Que, Jr. 1998. Oxygen activating nonheme iron enzymes. Curr Opin Chem Biol. 2:159-72.

65 Lavieu, G., F. Scarlatti, G. Sala, S. Carpentier, T. Levade, R. Ghidoni, J. Botti, and P. Codogno. 2006. Regulation of autophagy by sphingosine kinase 1 and its role in cell survival during nutrient starvation. Journal of Biological Chemistry. 281:8518-8527.

Le Gac, G., C. Mura, and C. Ferec. 2001. Complete scanning of the hereditary hemochromatosis gene (HFE) by use of denaturing HPLC. Clinical Chemistry. 47:1633-1640.

Lebron, J.A., M.J. Bennett, D.E. Vaughn, A.J. Chirino, P.M. Snow, G.A. Mintier, J.N. Feder, and P.J. Bjorkman. 1998. Crystal structure of the hemochromatosis protein HFE and characterization of its interaction with transferrin receptor. Cell. 93:111- 123.

Lebron, J.A., A.G. West, and P.J. Bjorkman. 1999. The hemochromatosis protein HFE competes with transferrin for binding to the transferrin receptor. Journal of Molecular Biology. 294:239-245.

Ledeen, R.W., and G.S. Wu. 2002. Ganglioside function in calcium homeostasis and signaling. Neurochemical Research. 27:637-647.

Ledesma, M.D. 2003. Raft disorganization leads to reduced plasmin activity in Alzheimer's disease brains. EMBO Rep. 4:1190-1196.

Ledesma, M.D., and C.G. Dotti. 2006. Amyloid excess in Alzheimer's disease: what is cholesterol to be blamed for? FEBS Lett. 580:5525-5532.

Lee, S.J., U. Liyanage, P.E. Bickel, W.M. Xia, P.T. Lansbury, and K.S. Kosik. 1998. A detergent-insoluble membrane compartment contains A beta in vivo. Nature Medicine. 4:730-734.

Lee, S.Y., and J.R. Connor. 2005. Regulation of Hfe by stress factors in BV-2 cells. Neurobiol Aging. 26:803-12.

66 Lee, S.Y., S.M. Patton, R.J. Henderson, and J.R. Connor. 2007. Consequences of expressing mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking endogenous HFE. FASEB J. 21:564-76.

Lehmann, D.J., M. Schuur, D.R. Warden, N. Hammond, O. Belbin, H. Kolsch, M.G. Lehmann, G.K. Wilcock, K. Brown, P.G. Kehoe, C.M. Morris, R. Barker, E. Coto, V. Alvarez, P. Deloukas, I. Mateo, R. Gwilliam, O. Combarros, A. Arias- Vasquez, Y.S. Aulchenko, M.A. Ikram, M.M. Breteler, C.M. van Duijn, A. Oulhaj, R. Heun, M. Cortina-Borja, K. Morgan, K. Robson, and A.D. Smith. 2012. Transferrin and HFE genes interact in Alzheimer's disease risk: the Epistasis Project. Neurobiology of Aging. 33.

Leoni, V., T. Masterman, U. Diczfalusy, G. De Luca, J. Hillert, and I. Bjorkhem. 2002. Changes in human plasma levels of the brain specific oxysterol 24S- hydroxycholesterol during progression of multiple sclerosis. Neurosci Lett. 331:163-6.

Li, H., S. Wetten, L. Li, P.L. St Jean, R. Upmanyu, L. Surh, D. Hosford, M.R. Barnes, J.D. Briley, M. Borrie, N. Coletta, R. Delisle, D. Dhalla, M.G. Ehm, H.H. Feldman, L. Fornazzari, S. Gauthier, N. Goodgame, D. Guzman, S. Hammond, P. Hollingworth, G.Y. Hsiung, J. Johnson, D.D. Kelly, R. Keren, A. Kertesz, K.S. King, S. Lovestone, I. Loy-English, P.M. Matthews, M.J. Owen, M. Plumpton, W. Pryse-Phillips, R.K. Prinjha, J.C. Richardson, A. Saunders, A.J. Slater, P.H. St George-Hyslop, S.W. Stinnett, J.E. Swartz, R.L. Taylor, J. Wherrett, J. Williams, D.P. Yarnall, R.A. Gibson, M.C. Irizarry, L.T. Middleton, and A.D. Roses. 2008. Candidate single-nucleotide polymorphisms from a genomewide association study of Alzheimer disease. Arch Neurol. 65:45-53.

Lin, M., L. Zhao, J. Fan, X.G. Lian, J.X. Ye, L. Wu, and H. Lin. 2012. Association between HFE polymorphisms and susceptibility to Alzheimer's disease: a meta- analysis of 22 studies including 4,365 cases and 8,652 controls. Molecular Biology Reports. 39:3089-3095.

67 Lingwood, D., and K. Simons. 2010. Lipid Rafts As a Membrane-Organizing Principle. Science. 327:46-50.

Liu, G., M.R. Garrett, P. Men, X. Zhu, G. Perry, and M.A. Smith. 2005. Nanoparticle and other metal chelation therapeutics in Alzheimer disease. Biochim Biophys Acta. 1741:246-52.

Liu, G., W. Huang, R.D. Moir, C.R. Vanderburg, B. Lai, Z. Peng, R.E. Tanzi, J.T. Rogers, and X. Huang. 2006. Metal exposure and Alzheimer's pathogenesis. J Struct Biol. 155:45-51.

Liu, Y., S.Y. Lee, E. Neely, W. Nandar, M. Moyo, Z. Simmons, and J.R. Connor. 2011. Mutant HFE H63D protein is associated with prolonged endoplasmic reticulum stress and increased neuronal vulnerability. J Biol Chem. 286:13161-70.

Lleo, A., R. Blesa, C. Angelopoulos, P. Pastor-Rubio, M. Villa, R. Oliva, and E. Bufill. 2002. Transferrin C2 allele, haemochromatosis gene mutations, and risk for Alzheimer's disease. J Neurol Neurosurg Psychiatry. 72:820-1.

Lloyd, R.V., P.M. Hanna, and R.P. Mason. 1997. The origin of the hydroxyl radical oxygen in the Fenton reaction. Free Radical Biology and Medicine. 22:885-888.

Loeffler, D.A., J.R. Connor, P.L. Juneau, B.S. Snyder, L. Kanaley, A.J. DeMaggio, H. Nguyen, C.M. Brickman, and P.A. LeWitt. 1995. Transferrin and iron in normal, Alzheimer's disease, and Parkinson's disease brain regions. J Neurochem. 65:710- 24.

Lopes, K.O., D.L. Sparks, and W.J. Streit. 2008. Microglial dystrophy in the aged and Alzheimer's disease brain is associated with ferritin immunoreactivity. Glia. 56:1048-1060.

Lukiw, W.J. 2006. Cholesterol and 24S-hydroxycholesterol trafficking in Alzheimer's disease. Expert Rev Neurother. 6:683-93.

Lutjohann, D. 2006. Cholesterol metabolism in the brain: importance of 24S- hydroxylation. Acta Neurol Scand Suppl. 185:33-42.

68 Lutjohann, D., O. Breuer, G. Ahlborg, I. Nennesmo, A. Siden, U. Diczfalusy, and I. Bjorkhem. 1996. Cholesterol homeostasis in human brain: evidence for an age- dependent flux of 24S-hydroxycholesterol from the brain into the circulation. Proc Natl Acad Sci U S A. 93:9799-804.

Lutjohann, D., A. Papassotiropoulos, I. Bjorkhem, S. Locatelli, M. Bagli, R.D. Oehring, U. Schlegel, F. Jessen, M.L. Rao, K. von Bergmann, and R. Heun. 2000. Plasma 24S-hydroxycholesterol (cerebrosterol) is increased in Alzheimer and vascular demented patients. J Lipid Res. 41:195-8.

Mahley, R.W. 1988. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science. 240:622-630.

Mahley, R.W., K.H. Weisgraber, and Y. Huang. 2006. Apolipoprotein E4: a causative factor and therapeutic target in neuropathology, including Alzheimer's disease. Proc. Natl Acad. Sci. USA. 103:5644-5651.

Mairuae, N., E.C. Hall Ii, P. Cheepsunthorn, S.Y. Lee, and J.R. Connor. 2010. The H63D HFE gene variant promotes activation of the intrinsic apoptotic pathway via mitochondria dysfunction following beta-amyloid peptide exposure. J Neurosci Res. 88:3079-89.

Manelli, A.M., W.B. Stine, L.J. Van Eldik, and M.J. LaDu. 2004. ApoE and Abeta1-42 interactions: effects of isoform and conformation on structure and function. J Mol Neurosci. 23:235-46.

Mantyh, P.W., J.R. Ghilardi, S. Rogers, E. Demaster, C.J. Allen, E.R. Stimson, and J.E. Maggio. 1993. Aluminum, Iron, and Zinc Ions Promote Aggregation of Physiological Concentrations of Beta-Amyloid Peptide. Journal of Neurochemistry. 61:1171-1174.

Marcus, J., and B. Popko. 2002. Galactolipids are molecular determinants of myelin development and axo-glial organization. Biochimica Et Biophysica Acta-General Subjects. 1573:406-413.

69 Markesbery, W.R. 1997. Oxidative stress hypothesis in Alzheimer's disease. Free Radic Biol Med. 23:134-47.

Marta, C.B., C.M. Taylor, S. Cheng, R.H. Quarles, R. Bansal, and S.E. Pfeiffer. 2004. Myelin associated glycoprotein cross-linking triggers its partitioning into lipid rafts, specific signaling events and cytoskeletal rearrangements in oligodendrocytes. Neuron Glia Biology. 1:35-46.

Martin, W.R., F.Q. Ye, and P.S. Allen. 1998. Increasing striatal iron content associated with normal aging. Mov Disord. 13:281-6.

Masliah, E., M. Mallory, M. Alford, R. DeTeresa, L.A. Hansen, D.W. McKeel, Jr., and J.C. Morris. 2001. Altered expression of synaptic proteins occurs early during progression of Alzheimer's disease. Neurology. 56:127-9.

Maynard, C.J., A.I. Bush, C.L. Masters, R. Cappai, and Q.X. Li. 2005. Metals and amyloid-beta in Alzheimer's disease. Int J Exp Pathol. 86:147-59.

Mcmillan, P.N., N.I. Williams, E.D. Day, and B. Kaufman. 1972. Isolation and Biochemical Characterization of 3 Subfractions of Myelin from Central Nervous- Tissue of Adult Rat. Journal of Neurochemistry. 19:1839-&.

Meaney, S., M. Hassan, A. Sakinis, D. Lutjohann, K. von Bergmann, A. Wennmalm, U. Diczfalusy, and I. Bjorkhem. 2001. Evidence that the major oxysterols in human circulation originate from distinct pools of cholesterol: a stable isotope study. Journal of Lipid Research. 42:70-78.

Merryweather-Clarke, A.T., J.J. Pointon, A.M. Jouanolle, J. Rochette, and K.J.H. Robson. 2000. Geography of HFE C282Y and H63D mutations. Genetic Testing. 4:183-198.

Mesulam, M.M., and M.A. Moran. 1987. Cholinesterases within Neurofibrillary Tangles Related to Age and Alzheimers-Disease. Annals of Neurology. 22:223-228.

70 Michikawa, M., Q.W. Fan, I. Isobe, and K. Yanagisawa. 2000. Apolipoprotein E exhibits isoform-specific promotion of lipid efflux from astrocytes and neurons in culture. J. Neurochem. 74:1008-1016.

Miller, R.H. 2002. Regulation of oligodendrocyte development in the vertebrate CNS. Progress in Neurobiology. 67:451-467.

Milton, W.J., S.W. Atlas, F.J. Lexa, P.D. Mozley, and R.E. Gur. 1991. Deep gray matter hypointensity patterns with aging in healthy adults: MR imaging at 1.5 T. Radiology. 181:715-9.

Mitchell, R.M., S.Y. Lee, W.T. Randazzo, Z. Simmons, and J.R. Connor. 2009a. Influence of HFE variants and cellular iron on monocyte chemoattractant protein- 1. J Neuroinflammation. 6:6.

Mitchell, R.M., S.Y. Lee, Z. Simmons, and J.R. Connor. 2009b. HFE polymorphisms affect cellular glutamate regulation. Neurobiol Aging.

Mitter, D., C. Reisinger, B. Hinz, S. Hollmann, S.V. Yelamanchili, S. Treiber-Held, T.G. Ohm, A. Herrmann, and G. Ahnert-Hilger. 2003. The synaptophysin/synaptobrevin interaction critically depends on the cholesterol content. Journal of Neurochemistry. 84:35-42.

Miyata, M., and J.D. Smith. 1996. Apolipoprotein E allele-specific antioxidant activity and effects on cytotoxicity by oxidative insults and beta-amyloid peptides. Nat Genet. 14:55-61.

Miyazaki, J.I., E. Appella, and K. Ozato. 1986. Intracellular-Transport Blockade Caused by Disruption of the Disulfide Bridge in the 3rd External Domain of Major Histocompatibility Complex Class-I Antigen. Proceedings of the National Academy of Sciences of the United States of America. 83:757-761.

Moalem, S., M.E. Percy, D.F. Andrews, T.P. Kruck, S. Wong, A.J. Dalton, P. Mehta, B. Fedor, and A.C. Warren. 2000. Are hereditary hemochromatosis mutations involved in Alzheimer disease? Am J Med Genet. 93:58-66.

71 Mohanty, J.G., H.D. Shukla, J.D. Williamson, L.J. Launer, S. Saxena, and J.M. Rifkind. 2010. Alterations in the red blood cell membrane proteome in alzheimer's subjects reflect disease-related changes and provide insight into altered cell morphology. Proteome Science. 8.

Molander-Melin, M., K. Blennow, N. Bogdanovic, B. Dellheden, J.E. Mansson, and P. Fredman. 2005. Structural membrane alterations in Alzheimer brains found to be associated with regional disease development; increased density of gangliosides GM1 and GM2 and loss of cholesterol in detergent-resistant membrane domains. Journal of Neurochemistry. 92:171-182.

Montine, T.J., W.R. Markesbery, W. Zackert, S.C. Sanchez, L.J. Roberts, and J.D. Morrow. 1999. The magnitude of brain lipid peroxidation correlates with the extent of degeneration but not with density of neuritic plaques or neurofibrillary tangles or with APOE genotype in Alzheimer's disease patients. American Journal of Pathology. 155:863-868.

Morgan, K. 2011. The three new pathways leading to Alzheimer's disease. Neuropathol Appl Neurobiol. 37:353-7.

Morishima-Kawashima, M., and Y. Ihara. 1998. The presence of amyloid beta-protein in the detergent-insoluble membrane compartment of human neuroblastoma cells. Biochemistry. 37:15247-15253.

Morishima-Kawashima, M., and Y. Ihara. 2002. Alzheimer's disease: beta-Amyloid protein and tau. J Neurosci Res. 70:392-401.

Muhlenhoff, U., S. Molik, J.R. Godoy, M.A. Uzarska, N. Richter, A. Seubert, Y. Zhang, J. Stubbe, F. Pierrel, E. Herrero, C.H. Lillig, and R. Lill. 2010. Cytosolic Monothiol Glutaredoxins Function in Intracellular Iron Sensing and Trafficking via Their Bound Iron-Sulfur Cluster. Cell Metabolism. 12:373-385.

Mullan, M., F. Crawford, K. Axelman, H. Houlden, L. Lilius, B. Winblad, and L. Lannfelt. 1992. A Pathogenic Mutation for Probable Alzheimers-Disease in the App Gene at the N-Terminus of Beta-Amyloid. Nature Genetics. 1:345-347.

72 Murrell, J.R., A.M. Hake, K.A. Quaid, M.R. Farlow, and B. Ghetti. 2000. Early-onset Alzheimer disease caused by a new mutation (V717L) in the amyloid precursor protein gene. Archives of Neurology. 57:885-887.

Mutter, J., J. Naumann, C. Sadaghiani, R. Schneider, and H. Walach. 2004. Alzheimer disease: mercury as pathogenetic factor and apolipoprotein E as a moderator. Neuro Endocrinol Lett. 25:331-9.

Nagy, Z., K.A. Jobst, M.M. Esiri, J.H. Morris, E.M. King, B. MacDonald, S. Litchfield, L. Barnetson, and A.D. Smith. 1996. Hippocampal pathology reflects memory deficit and brain imaging measurements in Alzheimer's disease: clinicopathologic correlations using three sets of pathologic diagnostic criteria. Dementia. 7:76-81.

Naj, A.C., G. Jun, G.W. Beecham, L.S. Wang, B.N. Vardarajan, J. Buros, P.J. Gallins, J.D. Buxbaum, G.P. Jarvik, P.K. Crane, E.B. Larson, T.D. Bird, B.F. Boeve, N.R. Graff-Radford, P.L. De Jager, D. Evans, J.A. Schneider, M.M. Carrasquillo, N. Ertekin-Taner, S.G. Younkin, C. Cruchaga, J.S. Kauwe, P. Nowotny, P. Kramer, J. Hardy, M.J. Huentelman, A.J. Myers, M.M. Barmada, F.Y. Demirci, C.T. Baldwin, R.C. Green, E. Rogaeva, P. St George-Hyslop, S.E. Arnold, R. Barber, T. Beach, E.H. Bigio, J.D. Bowen, A. Boxer, J.R. Burke, N.J. Cairns, C.S. Carlson, R.M. Carney, S.L. Carroll, H.C. Chui, D.G. Clark, J. Corneveaux, C.W. Cotman, J.L. Cummings, C. DeCarli, S.T. DeKosky, R. Diaz-Arrastia, M. Dick, D.W. Dickson, W.G. Ellis, K.M. Faber, K.B. Fallon, M.R. Farlow, S. Ferris, M.P. Frosch, D.R. Galasko, M. Ganguli, M. Gearing, D.H. Geschwind, B. Ghetti, J.R. Gilbert, S. Gilman, B. Giordani, J.D. Glass, J.H. Growdon, R.L. Hamilton, L.E. Harrell, E. Head, L.S. Honig, C.M. Hulette, B.T. Hyman, G.A. Jicha, L.W. Jin, N. Johnson, J. Karlawish, A. Karydas, J.A. Kaye, R. Kim, E.H. Koo, N.W. Kowall, J.J. Lah, A.I. Levey, A.P. Lieberman, O.L. Lopez, W.J. Mack, D.C. Marson, F. Martiniuk, D.C. Mash, E. Masliah, W.C. McCormick, S.M. McCurry, A.N. McDavid, A.C. McKee, M. Mesulam, B.L. Miller, et al. 2011. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease. Nat Genet. 43:436-41.

73 Nandar, W., and J.R. Connor. HFE gene variants affect iron in the brain. J Nutr. 141:729S-739S.

Netz, D.J.A., C.M. Stith, M. Stumpfig, G. Kopf, D. Vogel, H.M. Genau, J.L. Stodola, R. Lill, P.M.J. Burgers, and A.J. Pierik. 2012. Eukaryotic DNA polymerases require an iron-sulfur cluster for the formation of active complexes. Nature Chemical Biology. 8:125-132.

Nielsen, J.E., L.N. Jensen, and K. Krabbe. 1995. Hereditary haemochromatosis: a case of iron accumulation in the basal ganglia associated with a parkinsonian syndrome. J Neurol Neurosurg Psychiatry. 59:318-21.

Nilsberth, C., A. Westlind-Danielsson, C.B. Eckman, M.M. Condron, K. Axelman, C. Forsell, C. Stenh, J. Luthman, D.B. Teplow, S.G. Younkin, J. Naslund, and L. Lannfelt. 2001. The 'Arctic' APP mutation (E693G) causes Alzheimer's disease by enhanced A beta protofibril formation. Nature Neuroscience. 4:887-893.

Notkola, I.L., R. Sulkava, J. Pekkanen, T. Erkinjuntti, C. Ehnholm, P. Kivinen, J. Tuomilehto, and A. Nissinen. 1998. Serum total cholesterol, apolipoprotein E epsilon 4 allele, and Alzheimer's disease. Neuroepidemiology. 17:14-20.

Oettinghaus, B., M. Licci, L. Scorrano, and S. Frank. 2012. Less than perfect divorces: dysregulated mitochondrial fission and neurodegeneration. Acta Neuropathol. 123:189-203.

Okada, S. 1996. Iron-induced tissue damage and cancer: the role of reactive oxygen species-free radicals. Pathol Int. 46:311-32.

Okada, T., M. Wakabayashi, K. Ikeda, and K. Matsuzaki. 2007. Formation of toxic fibrils of Alzheimer's amyloid beta-protein-(1-40) by monosialoganglioside GM1, a neuronal membrane component. Journal of Molecular Biology. 371:481-489.

Oma, S., S. Mawatari, K. Saito, C. Wakana, Y. Tsuboi, T. Yamada, and T. Fujino. 2012. Changes in phospholipid composition of erythrocyte membrane in Alzheimer's disease. Dement Geriatr Cogn Dis Extra. 2:298-303.

74 Opazo, C., X.D. Huang, R.A. Cherny, R.D. Moir, A.E. Roher, A.R. White, R. Cappai, C.L. Masters, R.E. Tanzi, N.C. Inestrosa, and A.I. Bush. 2002. Metalloenzyme- like activity of Alzheimer's disease beta-amyloid - Cu-dependent catalytic conversion of dopamine, cholesterol, and biological reducing agents to neurotoxic H2O2. Journal of Biological Chemistry. 277:40302-40308.

Osborne, N.J., L.C. Gurrin, K.J. Allen, C.C. Constantine, M.B. Delatycki, C.E. McLaren, D.M. Gertig, G.J. Anderson, M.C. Southey, J.K. Olynyk, L.W. Powell, J.L. Hopper, G.G. Giles, and D.R. English. 2010. HFE C282Y Homozygotes Are at Increased Risk of Breast and Colorectal Cancer. Hepatology. 51:1311-1318.

Panzenboeck, U., Z. Balazs, A. Sovic, A. Hrzenjak, S. Levak-Frank, A. Wintersperger, E. Malle, and W. Sattler. 2002. ABCA1 and scavenger receptor class B, type I, are modulators of reverse sterol transport at an in Vitro blood-brain barrier constituted of porcine brain capillary endothelial cells. Journal of Biological Chemistry. 277:42781-42789.

Parkkila, S., A.K. Parkkila, A. Waheed, R.S. Britton, X.Y. Zhou, R.E. Fleming, S. Tomatsu, B.R. Bacon, and W.S. Sly. 2000. Cell surface expression of HFE protein in epithelial cells, macrophages, and monocytes. Haematologica. 85:340- 5.

Percy, M., S. Moalem, A. Garcia, M.J. Somerville, M. Hicks, D. Andrews, A. Azad, P. Schwarz, R. Beheshti Zavareh, R. Birkan, C. Choo, V. Chow, S. Dhaliwal, V. Duda, A.L. Kupferschmidt, K. Lam, D. Lightman, K. Machalek, W. Mar, F. Nguyen, P.J. Rytwinski, E. Svara, M. Tran, K. Wheeler, L. Yeung, K. Zanibbi, R. Zener, M. Ziraldo, and M. Freedman. 2008a. Involvement of ApoE E4 and H63D in sporadic Alzheimer's disease in a folate-supplemented Ontario population. J Alzheimers Dis. 14:69-84.

75 M. Ziraldo, and M. Freedman. 2008b. Involvement of ApoE E4 and H63D in sporadic Alzheimer's disease in a folate-supplemented Ontario population. J Alzheimers Dis. 14:69-84.

Perez, G.I., A. Jurisicova, T. Matikainen, T. Moriyama, M.R. Kim, Y. Takai, J.K. Pru, R.N. Kolesnick, and J.L. Tilly. 2005. A central role for ceramide in the age- related acceleration of apoptosis in the female germline. Faseb Journal. 19:860-+.

Perl, D.P. 2010. Neuropathology of Alzheimer's disease. Mt Sinai J Med. 77:32-42.

Perry, G., R.J. Castellani, M.A. Smith, P.L. Harris, Z. Kubat, K. Ghanbari, P.K. Jones, G. Cordone, M. Tabaton, B. Wolozin, and H. Ghanbari. 2003. Oxidative damage in the olfactory system in Alzheimer's disease. Acta Neuropathol. 106:552-6.

Perry, G., R. Friedman, G. Shaw, and V. Chau. 1987. Ubiquitin Is Detected in Neurofibrillary Tangles and Senile Plaque Neurites of Alzheimer-Disease Brains. Proceedings of the National Academy of Sciences of the United States of America. 84:3033-3036.

Perry, G., A.K. Raina, A. Nunomura, T. Wataya, L.M. Sayre, and M.A. Smith. 2000. How important is oxidative damage? Lessons from Alzheimer's disease. Free Radic Biol Med. 28:831-4.

Petersen, R.C., G.E. Smith, S.C. Waring, R.J. Ivnik, E.G. Tangalos, and E. Kokmen. 1999. Mild cognitive impairment: clinical characterization and outcome. Arch Neurol. 56:303-8.

Petro, K.A., and C. Schengrund. 2007. Disruption of lipid rafts affects ganglioside composition and distribution in N2a cells. Journal of Neurochemistry. 102:234- 234.

Pettegrew, J.W., K. Panchalingam, R.L. Hamilton, and R.J. McClure. 2001. Brain membrane phospholipid alterations in Alzheimer's disease. Neurochemical Research. 26:771-782.

76 Pfrieger, F.W. 2003a. Outsourcing in the brain: do neurons depend on cholesterol delivery by astrocytes? Bioessays. 25:72-78.

Pfrieger, F.W. 2003b. Role of cholesterol in synapse formation and function. Biochimica Et Biophysica Acta-Biomembranes. 1610:271-280.

Pitas, R.E., J.K. Boyles, S.H. Lee, D. Foss, and R.W. Mahley. 1987a. Astrocytes synthesize apolipoprotein E and metabolize apolipoprotein E-containing lipoproteins. Biochim. Biophys. Acta. 917:148-161.

Pitas, R.E., J.K. Boyles, S.H. Lee, D. Foss, and R.W. Mahley. 1987b. Astrocytes synthesize apolipoprotein E and metabolize apolipoprotein E-containing lipoproteins. Biochim Biophys Acta. 917:148-61.

Pitas, R.E., J.K. Boyles, S.H. Lee, D. Hui, and K.H. Weisgraber. 1987c. Lipoproteins and their receptors in the central nervous system. Characterization of the lipoproteins in cerebrospinal fluid and identification of apolipoprotein B,E(LDL) receptors in the brain. J Biol Chem. 262:14352-60.

Plotz, E.J., J.J. Kabara, M.E. Davis, G.V. LeRoy, and R.G. Gould. 1968. Studies on the synthesis of cholesterol in the brain of the human fetus. Am J Obstet Gynecol. 101:534-8.

Potkin, S.G., G. Guffanti, A. Lakatos, J.A. Turner, F. Kruggel, J.H. Fallon, A.J. Saykin, A. Orro, S. Lupoli, E. Salvi, M. Weiner, and F. Macciardi. 2009. Hippocampal atrophy as a quantitative trait in a genome-wide association study identifying novel susceptibility genes for Alzheimer's disease. PLoS One. 4:e6501.

Pratico, D., C.M. Clark, F. Liun, V.Y.M. Lee, and J.Q. Trojanowski. 2002. Increase of brain oxidative stress in mild cognitive impairment - A possible predictor of Alzheimer disease. Archives of Neurology. 59:972-976.

Pulliam, J.F., C.D. Jennings, R.J. Kryscio, D.G. Davis, D. Wilson, T.J. Montine, F.A. Schmitt, and W.R. Markesbery. 2003. Association of HFE mutations with

77 neurodegeneration and oxidative stress in Alzheimer's disease and correlation with APOE. Am J Med Genet B Neuropsychiatr Genet. 119B:48-53.

Querfurth, H.W., and F.M. LaFerla. 2010. Alzheimer's disease. N Engl J Med. 362:329- 44.

Rahman, A., S. Akterin, A. Flores-Morales, M. Crisby, M. Kivipelto, M. Schultzberg, and A. Cedazo-Minguez. 2005. High cholesterol diet induces tau hyperphosphorylation in apolipoprotein E deficient mice. FEBS Lett. 579:6411-6.

Rapp, A., B. Gmeiner, and M. Huttinger. 2006. Implication of apoE isoforms in cholesterol metabolism by primary rat hippocampal neurons and astrocytes. Biochimie. 88:473-483.

Reddy, P.H. 2011. Abnormal tau, mitochondrial dysfunction, impaired axonal transport of mitochondria, and synaptic deprivation in Alzheimer's disease. Brain Research. 1415:136-148.

Reiman, E.M., J.A. Webster, A.J. Myers, J. Hardy, T. Dunckley, V.L. Zismann, K.D. Joshipura, J.V. Pearson, D. Hu-Lince, M.J. Huentelman, D.W. Craig, K.D. Coon, W.S. Liang, R.H. Herbert, T. Beach, K.C. Rohrer, A.S. Zhao, D. Leung, L. Bryden, L. Marlowe, M. Kaleem, D. Mastroeni, A. Grover, C.B. Heward, R. Ravid, J. Rogers, M.L. Hutton, S. Melquist, R.C. Petersen, G.E. Alexander, R.J. Caselli, W. Kukull, A. Papassotiropoulos, and D.A. Stephan. 2007. GAB2 alleles modify Alzheimer's risk in APOE epsilon4 carriers. Neuron. 54:713-20.

Richardson, D.R., D.J.R. Lane, E.M. Becker, M.L.H. Huang, M. Whitnall, Y.S. Rahmanto, A.D. Sheftel, and P. Ponka. 2010. Mitochondrial iron trafficking and the integration of iron metabolism between the mitochondrion and cytosol. Proceedings of the National Academy of Sciences of the United States of America. 107:10775-10782.

Riddell, D.R. 2008. Impact of apolipoprotein E (apoE) polymorphism on brain apoE levels. J. Neurosci. 28:11445-11453.

78 Riegert, P., S. Gilfillan, I. Nanda, M. Schmid, and S. Bahram. 1998. The mouse HFE gene. Immunogenetics. 47:174-177.

Roberts, B.R., T.M. Ryan, A.I. Bush, C.L. Masters, and J.A. Duce. The role of metallobiology and amyloid-beta peptides in Alzheimer's disease. J Neurochem. 120 Suppl 1:149-66.

Robson, K.J., D.J. Lehmann, V.L. Wimhurst, K.J. Livesey, M. Combrinck, A.T. Merryweather-Clarke, D.R. Warden, and A.D. Smith. 2004. Synergy between the C2 allele of transferrin and the C282Y allele of the haemochromatosis gene (HFE) as risk factors for developing Alzheimer's disease. J Med Genet. 41:261-5.

Rogers, J.T., J.D. Randall, C.M. Cahill, P.S. Eder, X.D. Huang, H. Gunshin, L. Leiter, J. McPhee, S.S. Sarang, T. Utsuki, N.H. Greig, D.K. Lahiri, R.E. Tanzi, A.I. Bush, T. Giordano, and S.R. Gullans. 2002. An iron-responsive element type II in the 5 '-untranslated region of the Alzheimer's amyloid precursor protein transcript. Journal of Biological Chemistry. 277:45518-45528.

Roher, A.E., Y.M. Kuo, K.M. Kokjohn, M.R. Emmerling, and S. Gracon. 1999. Amyloid and lipids in the pathology of Alzheimer disease. Amyloid. 6:136-45.

Rohrlich, P.S., N. Fazilleau, F. Ginhoux, H. Firat, F. Michel, M. Cochet, N. Laham, M.P. Roth, S. Pascolo, F. Nato, H. Coppin, P. Charneau, O. Danos, O. Acuto, R. Ehrlich, J. Kanellopoulos, and F.A. Lemonnier. 2005. Direct recognition by alpha beta cytolytic T cells of Hfe, a MHC class Ib molecule without antigen-presenting function. Proceedings of the National Academy of Sciences of the United States of America. 102:12855-12860.

Rottkamp, C.A., A. Nunomura, A.K. Raina, L.M. Sayre, G. Perry, and M.A. Smith. 2000. Oxidative stress, antioxidants, and Alzheimer disease. Alzheimer Dis Assoc Disord. 14 Suppl 1:S62-6.

Rottkamp, C.A., A.K. Raina, X. Zhu, E. Gaier, A.I. Bush, C.S. Atwood, M. Chevion, G. Perry, and M.A. Smith. 2001a. Redox-active iron mediates amyloid-beta toxicity. Free Radic Biol Med. 30:447-50.

79 Rottkamp, C.A., A.K. Raina, X.W. Zhu, E. Gaier, A.I. Bush, C.S. Atwood, M. Chevion, G. Perry, and M.A. Smith. 2001b. Redox-active iron mediates amyloid-beta toxicity. Free Radical Biology and Medicine. 30:447-450.

Rouault, T.A., and W.H. Tong. 2005. Iron-sulphur cluster biogenesis and mitochondrial iron homeostasis. Nature Reviews Molecular Cell Biology. 6:345-351.

Ryland, L.K., T.E. Fox, X. Liu, T.P. Loughran, and M. Kester. 2011. Dysregulation of sphingolipid metabolism in cancer. Cancer Biology & Therapy. 11:138-149.

S. Moalem, M.E. Percy, D.F. Andrews, T.P. Kruck, S. Wong, A.J. Dalton, P. Mehta, B. Fedor, and A.C. Warren. 2000. Erratum: Are hereditary hemochromatosis mutations involved in alzheimer disease? American Journal of Medical Genetics. 95:189.

Sabolovic, D., M. Roudier, M. Boynard, C. Pautou, C. Sestier, B. Fertil, D. Geldwerth, J. Roger, J.N. Pons, A. Amri, and A. Halbreich. 1997. Membrane modifications of red blood cells in Alzheimer's disease. Journals of Gerontology Series a- Biological Sciences and Medical Sciences. 52:B217-B220.

Saher, G., B. Brugger, C. Lappe-Siefke, W. Mobius, R. Tozawa, M.C. Wehr, F. Wieland, S. Ishibashi, and K.A. Nave. 2005. High cholesterol level is essential for myelin membrane growth. Nat Neurosci. 8:468-75.

Saito, M., E.P. Benson, M. Saito, and A. Rosenberg. 1987. Metabolism of Cholesterol and Triacylglycerol in Cultured Chick Neuronal Cells, Glial-Cells, and Fibroblasts - Accumulation of Esterified Cholesterol in Serum-Free Culture. Journal of Neuroscience Research. 18:319-325.

Salter-Cid, L., A. Brunmark, P.A. Peterson, and Y. Yang. 2000. The major histocompatibility complex-encoded class I-like HFE abrogates endocytosis of transferrin receptor by inducing receptor phosphorylation. Genes and Immunity. 1:409-417.

80 Sampietro, M., L. Caputo, A. Casatta, M. Meregalli, A. Pellagatti, J. Tagliabue, G. Annoni, and C. Vergani. 2001. The hemochromatosis gene affects the age of onset of sporadic Alzheimer's disease. Neurobiol Aging. 22:563-8.

Satoi, H., H. Tomimoto, R. Ohtani, T. Kitano, T. Kondo, M. Watanabe, N. Oka, I. Akiguchi, S. Furuya, Y. Hirabayashi, and T. Okazaki. 2005. Astroglial expression of ceramide in Alzheimer's disease brains: A role during neuronal apoptosis. Neuroscience. 130:657-666.

Saunders, A.M., W.J. Strittmatter, D. Schmechel, P.H. George-Hyslop, M.A. Pericak- Vance, S.H. Joo, B.L. Rosi, J.F. Gusella, D.R. Crapper-MacLachlan, M.J. Alberts, and et al. 1993. Association of apolipoprotein E allele epsilon 4 with late-onset familial and sporadic Alzheimer's disease. Neurology. 43:1467-72.

Sayre, L.M., G. Perry, C.S. Atwood, and M.A. Smith. 2000. The role of metals in neurodegenerative diseases. Cell Mol Biol (Noisy-le-grand). 46:731-41.

Schengrund, C.L. 2010. Lipid rafts: Keys to neurodegeneration. Brain Research Bulletin. 82:7-17.

Schmechel, D.E. 1993. Increased amyloid [beta]-peptide deposition in cerebral cortex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc. Natl Acad. Sci. USA. 90:9649-9653.

Schreurs, B.G. 2010. The effects of cholesterol on learning and memory. Neurosci Biobehav Rev. 34:1366-79.

Schubert, D., and M. Chevion. 1995. The Role of Iron in Beta-Amyloid Toxicity. Biochemical and Biophysical Research Communications. 216:702-707.

Selkoe, D.J. 1996. Amyloid beta-protein and the genetics of Alzheimer's disease. J Biol Chem. 271:18295-8.

Selkoe, D.J. 2002a. Alzheimer's disease is a synaptic failure. Science. 298:789-91.

Selkoe, D.J. 2002b. Alzheimer's disease is a synaptic failure. Science. 298:789-791.

81 Selkoe, D.J. 2005. Defining molecular targets to prevent Alzheimer disease. Arch Neurol. 62:192-5.

Seshadri, S., A.L. Fitzpatrick, M.A. Ikram, A.L. DeStefano, V. Gudnason, M. Boada, J.C. Bis, A.V. Smith, M.M. Carassquillo, J.C. Lambert, D. Harold, E.M. Schrijvers, R. Ramirez-Lorca, S. Debette, W.T. Longstreth, Jr., A.C. Janssens, V.S. Pankratz, J.F. Dartigues, P. Hollingworth, T. Aspelund, I. Hernandez, A. Beiser, L.H. Kuller, P.J. Koudstaal, D.W. Dickson, C. Tzourio, R. Abraham, C. Antunez, Y. Du, J.I. Rotter, Y.S. Aulchenko, T.B. Harris, R.C. Petersen, C. Berr, M.J. Owen, J. Lopez-Arrieta, B.N. Varadarajan, J.T. Becker, F. Rivadeneira, M.A. Nalls, N.R. Graff-Radford, D. Campion, S. Auerbach, K. Rice, A. Hofman, P.V. Jonsson, H. Schmidt, M. Lathrop, T.H. Mosley, R. Au, B.M. Psaty, A.G. Uitterlinden, L.A. Farrer, T. Lumley, A. Ruiz, J. Williams, P. Amouyel, S.G. Younkin, P.A. Wolf, L.J. Launer, O.L. Lopez, C.M. van Duijn, and M.M. Breteler. 2010. Genome- wide analysis of genetic loci associated with Alzheimer disease. JAMA. 303:1832- 40.

Shah, R.S., H.G. Lee, X.W. Zhu, G. Perry, M.A. Smith, and R.J. Castellani. 2008. Current approaches in the treatment of Alzheimer's disease. Biomedicine & Pharmacotherapy. 62:199-207.

Shaheen, N.J., L.M. Silverman, T. Keku, L.B. Lawrence, E.M. Rohlfs, C.F. Martin, J. Galanko, and R.S. Sandler. 2003. Association between hemochromatosis (HFE) gene mutation carrier status and the risk of colon cancer. Journal of the National Cancer Institute. 95:154-159.

Shobab, L.A., G.Y.R. Hsiung, and H.H. Feldman. 2005. Cholesterol in Alzheimer's disease. Lancet Neurol. 4:841-852.

Simon, M., M. Bourel, R. Fauchet, and B. Genetet. 1976. Association of HLA-A3 and HLA-B14 antigens with idiopathic haemochromatosis. Gut. 17:332-4.

Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nature Reviews Molecular Cell Biology. 1:31-39.

82 Simpson, M.A., H. Cross, C. Proukakis, D.A. Priestman, D.C.A. Neville, G. Reinkensmeier, H. Wang, M. Wiznitzer, K. Gurtz, A. Verganelaki, A. Pryde, M.A. Patton, R.A. Dwek, T.D. Butters, F.M. Platt, and A.H. Crosby. 2004. Infantile-onset symptomatic epilepsy syndrome caused by a homozygous loss-of- function mutation of GM3 synthase. Nature Genetics. 36:1225-1229.

Sjogren, M., M. Mielke, D. Gustafson, P. Zandi, and I. Skoog. 2006. Cholesterol and Alzheimer's disease--is there a relation? Mech Ageing Dev. 127:138-47.

Skinner, E.R., C. Watt, J.A.O. Besson, and P.V. Best. 1989. Lipid-Composition of Different Regions of the Brain in Patients with Alzheimers-Disease. Biochemical Society Transactions. 17:213-214.

Smith, M.A. 2006. Oxidative stress and iron imbalance in Alzheimer disease: how rust became the fuss! J Alzheimers Dis. 9:305-8.

Snider, A.J., K.A.O. Gandy, and L.M. Obeid. 2010. Sphingosine kinase: Role in regulation of bioactive sphingolipid mediators in inflammation. Biochimie. 92:707-715.

Soderberg, M., C. Edlund, I. Alafuzoff, K. Kristensson, and G. Dallner. 1992. Lipid- Composition in Different Regions of the Brain in Alzheimers-Disease Senile Dementia of Alzheimers Type. Journal of Neurochemistry. 59:1646-1653.

Spiegel, S., and S. Milstien. 2002. Sphingosine 1-phosphate, a key cell signaling molecule. Journal of Biological Chemistry. 277:25851-25854.

Srivastava, R.A., N. Srivastava, M. Averna, R.C. Lin, K.S. Korach, D.B. Lubahn, and G. Schonfeld. 1997. Estrogen up-regulates apolipoprotein E (ApoE) gene expression by increasing ApoE mRNA in the translating pool via the estrogen receptor alpha- mediated pathway. J Biol Chem. 272:33360-6.

Stetzkowski-Marden, F., M. Recouvreur, G. Camus, A. Cartaud, S. Marchand, and J. Cartaud. 2006. Rafts are required for acetylcholine receptor clustering. Journal of Molecular Neuroscience. 30:37-38.

83 Stone, D.J., I. Rozovsky, T.E. Morgan, C.P. Anderson, H. Hajian, and C.E. Finch. 1997. Astrocytes and microglia respond to estrogen with increased apoE mRNA in vivo and in vitro. Exp Neurol. 143:313-8.

Stratman, N.C., C.K. Castle, B.M. Taylor, D.E. Epps, G.W. Melchior, and D.B. Carter. 2005. Isoform-specific interactions of human apolipoprotein E to an intermediate conformation of human Alzheimer amyloid-beta peptide. Chem Phys Lipids. 137:52-61.

Strittmater, W.J. 1994. Isoform-specific interactions of apolipoprotein E with microtubule associated protein tau: Implications for Alzheimer disease. Proc. Natl. Acad Sci. U. S. A. 91:11183-11186.

Strittmatter, W.J., A.M. Saunders, D. Schmechel, M. Pericak-Vance, J. Enghild, G.S. Salvesen, and A.D. Roses. 1993a. Apolipoprotein E: high-avidity binding to beta- amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A. 90:1977-81.

Strittmatter, W.J., K.H. Weisgraber, D.Y. Huang, L.M. Dong, G.S. Salvesen, M. Pericak- Vance, D. Schmechel, A.M. Saunders, D. Goldgaber, and A.D. Roses. 1993b. Binding of human apolipoprotein E to synthetic amyloid beta peptide: isoform- specific effects and implications for late-onset Alzheimer disease. Proc Natl Acad Sci U S A. 90:8098-102.

Suzuki, N., T.T. Cheung, X.D. Cai, A. Odaka, L. Otvos, C. Eckman, T.E. Golde, and S.G. Younkin. 1994. An Increased Percentage of Long Amyloid-Beta Protein Secreted by Familial Amyloid-Beta Protein-Precursor (Beta-App(717)) Mutants. Science. 264:1336-1340.

Suzuki, S., K. Kiyosue, S. Hazama, A. Ogura, M. Kashihara, T. Hara, H. Koshimizu, and M. Kojima. 2007. Brain-derived neurotrophic factor regulates cholesterol metabolism for synapse development. J Neurosci. 27:6417-27.

Suzuki, S., T. Numakawa, K. Shimazu, H. Koshimizu, T. Hara, H. Hatanaka, L. Mei, B. Lu, and M. Kojima. 2004. BDNF-induced recruitment of TrkB receptor into

84 neuronal lipid rafts: roles in synaptic modulation. Journal of Cell Biology. 167:1205-1215.

Svennerholm, L., and C.G. Gottfries. 1994. Membrane-Lipids, Selectively Diminished in Alzheimer Brains, Suggest Synapse Loss as a Primary Event in Early-Onset Form (Type-I) and Demyelination in Late-Onset Form (Type-Ii). Journal of Neurochemistry. 62:1039-1047.

Syrjakoski, K., H. Fredriksson, T. Ikonen, T. Kuukasjarvi, V. Autio, M.P. Matikainen, T.L.J. Tammela, P.A. Koivistol, and J. Schleutker. 2006. Hemochromatosis gene mutations among Finnish male breast and prostate cancer patients. International Journal of Cancer. 118:518-520.

Sze, C.I., J.C. Troncoso, C. Kawas, P. Mouton, D.L. Price, and L.J. Martin. 1997. Loss of the presynaptic vesicle protein synaptophysin in hippocampus correlates with cognitive decline in Alzheimer disease. J Neuropathol Exp Neurol. 56:933-44.

Takahashi, S., I. Takahashi, H. Sato, Y. Kubota, S. Yoshida, and Y. Muramatsu. 2001. Age-related changes in the concentrations of major and trace elements in the brain of rats and mice. Biological Trace Element Research. 80:145-158.

Takamori, S., M. Holt, K. Stenius, E.A. Lemke, M. Gronborg, D. Riedel, H. Urlaub, S. Schenck, B. Brugger, P. Ringler, S.A. Muller, B. Rammner, F. Grater, J.S. Hub, B.L. De Groot, G. Mieskes, Y. Moriyama, J. Klingauf, H. Grubmuller, J. Heuser, F. Wieland, and R. Jahn. 2006. Molecular anatomy of a trafficking organelle. Cell. 127:831-846.

Terry, R.D. 1963. The Fine Structure of Neurofibrillary Tangles in Alzheimer's Disease. J Neuropathol Exp Neurol. 22:629-42.

Terry, R.D., E. Masliah, D.P. Salmon, N. Butters, R. DeTeresa, R. Hill, L.A. Hansen, and R. Katzman. 1991. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Ann Neurol. 30:572- 80.

85 Tesseur, I. 2000. Expression of human apolipoprotein E4 in neurons causes hyperphosphorylation of protein tau in the brains of transgenic mice. Am. J. Pathol. 156:951-964.

Teter, B., and C.E. Finch. 2004. Caliblan's heritance and the genetics of neuronal aging. Trends in Neurosciences. 27:627-632.

Thiele, C., M.J. Hannah, F. Fahrenholz, and W.B. Huttner. 2000. Cholesterol binds to synaptophysin and is required for biogenesis of synaptic vesicles. Nat Cell Biol. 2:42-9.

Thinakaran, G. 1999. The role of presenilins in Alzheimer's disease. J Clin Invest. 104:1321-7.

Todorich, B.M., and J.R. Connor. 2004. Redox metals in Alzheimer's disease. Ann N Y Acad Sci. 1012:171-8.

Tong, W.H., and T.A. Rouault. 2006. Functions of mitochondrial ISCU and cytosolic ISCU in mammalian iron-sulfur cluster biogenesis and iron homeostasis. Cell Metabolism. 3:199-210.

Tremollieres, F.A., J.M. Pouilles, C. Cauneille, and C. Ribot. 1999. Coronary heart disease risk factors and menopause: A study in 1684 French women. Atherosclerosis. 142:415-423.

Trommer, B.L., C. Shah, S.H. Yun, G. Gamkrelidze, E.S. Pasternak, W.B. Stine, A. Manelli, P. Sullivan, J.F. Pasternak, and M.J. LaDu. 2005. ApoE isoform-specific effects on LTP: blockade by oligomeric amyloid-beta1-42. Neurobiol Dis. 18:75- 82.

Tsui-Pierchala, B.A., M. Encinas, J. Milbrandt, and E.M. Johnson. 2002. Lipid rafts in neuronal signaling and function. Trends in Neurosciences. 25:412-417.

Uttara, B., A.V. Singh, P. Zamboni, and R.T. Mahajan. 2009. Oxidative stress and neurodegenerative diseases: a review of upstream and downstream antioxidant therapeutic options. Curr Neuropharmacol. 7:65-74.

86 Valenza, M., D. Rigamonti, D. Goffredo, C. Zuccato, S. Fenu, L. Jamot, A. Strand, A. Tarditi, B. Woodman, M. Racchi, C. Mariotti, S. Di Donato, A. Corsini, G. Bates, R. Pruss, J.M. Olson, S. Sipione, M. Tartari, and E. Cattaneo. 2005. Dysfunction of the cholesterol biosynthetic pathway in Huntington's disease. Journal of Neuroscience. 25:9932-9939.

Vance, J.E. 2006. Lipid imbalance in the neurological disorder, Niemann-Pick C disease. Febs Letters. 580:5518-5524.

Vaya, J., and H.M. Schipper. 2007. Oxysterols, cholesterol homeostasis, and Alzheimer disease. J. Neurochem. 102:1727-1737.

Veatch, S.L., and S.L. Keller. 2005. Seeing spots: Complex phase behavior in simple membranes. Biochimica Et Biophysica Acta-Molecular Cell Research. 1746:172- 185.

Venable, M.E., J.Y. Lee, M.J. Smyth, A. Bielawska, and L.M. Obeid. 1995. Role of ceramide in cellular senescence. Journal of Biological Chemistry. 270:30701- 30708.

Vetrivel, K.S., H.P. Cheng, S.H. Kim, Y. Chen, N.Y. Barnes, A.T. Parent, S.S. Sisodia, and G. Thinakaran. 2005. Spatial segregation of gamma-secretase and substrates in distinct membrane domains. Journal of Biological Chemistry. 280:25892- 25900.

Waelsch, H., W.M. Sperry, and V.A. Stoyanoff. 1940. A study of the synthesis and deposition of lipids in brain and other tissues with deuterium as an indicator. Journal of Biological Chemistry. 135:291-296.

Waheed, A., J.H. Grubb, X.Y. Zhou, S. Tomatsu, R.E. Fleming, M.E. Costaldi, R.S. Britton, B.R. Bacon, and W.S. Sly. 2002. Regulation of transferrin-mediated iron uptake by HFE, the protein defective in hereditary hemochromatosis. Proc Natl Acad Sci U S A. 99:3117-22.

87 Waheed, A., S. Parkkila, J. Saarnio, R.E. Fleming, X.Y. Zhou, S. Tomatsu, R.S. Britton, B.R. Bacon, and W.S. Sly. 1999. Association of HFE protein with transferrin receptor in crypt enterocytes of human duodenum. Proc Natl Acad Sci U S A. 96:1579-84.

Waheed, A., S. Parkkila, X.Y. Zhou, S. Tomatsu, Z. Tsuchihashi, J.N. Feder, R.C. Schatzman, R.S. Britton, B.R. Bacon, and W.S. Sly. 1997. Hereditary hemochromatosis: Effects of C282Y and H63D mutations on association with beta(2)-microglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proceedings of the National Academy of Sciences of the United States of America. 94:12384-12389.

Walsh, D.M., and D.J. Selkoe. 2004. Deciphering the molecular basis of memory failure in Alzheimer's disease. Neuron. 44:181-93.

Wang, X.S., S. Lee, Z. Simmons, P. Boyer, K. Scott, W. Liu, and J. Connor. 2004. Increased incidence of the Hfe mutation in amyotrophic lateral sclerosis and related cellular consequences. J Neurol Sci. 227:27-33.

Wellington, C.L. 2004. Cholesterol at the crossroads: Alzheimer's disease and lipid metabolism. Clinical Genetics. 66:1-16.

Willmann, R., S. Pun, L. Stallmach, G. Sadasivam, A.F. Santos, P. Caroni, and C. Fuhrer. 2006. Cholesterol and lipid microdomains stabilize the postsynapse at the neuromuscular junction. Embo Journal. 25:4050-4060.

Wingo, T.S., J.J. Lah, A.I. Levey, and D.J. Cutler. 2012. Autosomal recessive causes likely in early-onset Alzheimer disease. Arch Neurol. 69:59-64.

Wolozin, B. 2003. Cyp46 (24S-cholesterol hydroxylase): a genetic risk factor for Alzheimer disease. Arch Neurol. 60:16-8.

Wolozin, B. 2004. Cholesterol, statins and dementia. Current Opinion in Lipidology. 15:667-672.

88 Wu, G.S., Z.H. Lu, A.G. Obukhov, M.C. Nowycky, and R.W. Ledeen. 2007. Induction of calcium influx through TRPC5 channels by cross-linking of GM1 ganglioside associated with alpha 5 beta 1 integrin initiates neurite outgrowth. Journal of Neuroscience. 27:7447-7458.

Yamamoto, N., Y. Fukata, M. Fukata, and K. Yanagisawa. 2007. GM1-ganglioside- induced Abeta assembly on synaptic membranes of cultured neurons. Biochim Biophys Acta. 1768:1128-37.

Yanagisawa, K. 2005. GM1 ganglioside and the seeding of amyloid in Alzheimer's disease: endogenous seed for Alzheimer amyloid. Neuroscientist. 11:250-60.

Yankner, B.A. 1996. New clues to Alzheimer's disease: unraveling the roles of amyloid and tau. Nat Med. 2:850-2.

Yankner, B.A., T. Lu, and P. Loerch. 2008. The aging brain. Annual Review of Pathology-Mechanisms of Disease. 3:41-66.

Yao, P.J., M. Zhu, E.I. Pyun, A.I. Brooks, S. Therianos, V.E. Meyers, and P.D. Coleman. 2003. Defects in expression of genes related to synaptic vesicle trafficking in frontal cortex of Alzheimer's disease. Neurobiology of Disease. 12:97-109.

Zerbinatti, C.V., and G. Bu. 2005. LRP and Alzheimer's disease. Rev Neurosci. 16:123- 35.

Zha, Q., Y. Ruan, T. Hartmann, K. Beyreuther, and D. Zhang. 2004. GM1 ganglioside regulates the proteolysis of amyloid precursor protein. Molecular Psychiatry. 9:946-952.

89 Chapter 2

H63D variant of HFE is associated with disruption of cholesterol metabolism, brain atrophy, and cognitive impairment: implications for Alzheimer Disease

ABSTRACT

The H63D variant of the hemochromatosis (HFE) gene, associated with deregulation of iron homeostasis, has been implicated as a risk factor for Alzheimer disease (AD). Apolipoprotein epsilon 4 (APOE4), a known risk factor for AD, is associated with disruption of cholesterol metabolism. Recent studies showed that individuals with both the APOE4 allele and the H63D variant had a much greater risk and earlier onset of AD than individuals carrying either the APOE4 allele or H63D variant of HFE alone. Therefore, we tested the hypothesis that H63D-HFE is associated with development of cognitive impairment by disrupting cholesterol metabolism thereby contributing to increased neurodegeneration, and increased memory deficits. The effect(s) of H63D-HFE on expression of cholesterol was investigated using SH-SY5Y human neuroblastoma cells transfected to stably express either wild type (WT-) or H63D-HFE. The ~50% reduction in cholesterol content in cells expressing H63D-HFE was accompanied by a significant decrease in expression of 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMGCoAR), and a significant increase in expression of cholesterol hydroxylase (CYP46A1). Consistent with in vitro studies, H67D-HFE (homologous to human H63D- HFE) knock-in mice, showed a genotype and age dependent decline in brain cholesterol relative to WT-HFE controls and similar changes in expression of proteins regulating cholesterol metabolism. The reduction in cholesterol in brains of H67D-HFE mice was accompanied by a significant decrease in expression of synapse and myelin management proteins, significant increase in caspase-3 expression, and cortical and hippocampal neurodegeneration in the brains of 18 and 24 month old mice. Performance on learning

90 and memory tasks (novel object recognition and Barnes maze) indicated that H67D-HFE mice had poorer cognitive function than WT-HFE mice. Magnetic resonance imaging of 24 month-old mice showed a greater reduction in volume of brains in H67D-HFE mice. Combined these data indicate that the H63D/H67D variant contributes to the disease process in AD by altering iron and cholesterol metabolism, which is accompanied by neuronal loss, brain atrophy, and memory impairment which can promote earlier onset of AD seen in carriers of H63D.

Introduction

Alzheimer disease (AD), the most common form of dementia, affects more than 5.3 million people in the United States. AD is the seventh cause of death overall and the fifth cause of death for those over 65. It is estimated that about half of those aged 85 and over have significant cognitive impairment. While the mechanism(s) that results in the pathogenesis of AD are not completely understood, disruption of cholesterol homeostasis (Shobab et al., 2005; Sjogren et al., 2006) and metal-induced oxidative damage (Arosio and Levi, 2002; Connor and Lee, 2006; Fuhrman et al., 1994; Uttara et al., 2009) have been consistently reported in AD. Despite advances made in understanding potential effects of each of these pathways, little attention has been given to the influence of alterations in each of these pathways on the etiology of this disorder. The observation that characteristics associated with AD can be induced by either cholesterol or iron dysregulation, supports the hypothesis that disruption of one pathway could influence the other. Despite the relevance of disruptions in either the iron or cholesterol pathway to development of AD, the interactions of these pathways in AD have not been explored.

Disruption of cholesterol homeostasis has been strongly implicated in AD with the strongest risk factor identified thus far for late onset of AD being the presence of the apolipoprotein E Ɛ4 allele of the apolipoprotein E (ApoE) gene. The ApoE4 allele which encodes the ApoE4 protein needed for cholesterol transport is found in more than 40% of those affected with AD. In contrast, the apolipoprotein E Ɛ2 allele is associated with

91 reduced risk of AD while the apolipoprotein E Ɛ3 allele is considered the norm. The lack of functional ApoE in brains isolated from mice expressing ApoE4 is accompanied by reduced levels of neuronal cholesterol (Gong et al., 2002) presumably due to the decreased ability of ApoE4 to deliver cholesterol to them (Bu, 2009). Variations observed in phenotype (onset, progression, and severity) of disease in ApoE4 allele carriers suggest the presence of additional genetic modifiers of AD in the population.

One such possible modifier is HFE (high iron protein) which is known to regulate iron uptake (for a recent review see (Connor and Lee, 2006; Nandar and Connor). Iron accumulation has been found in the vicinity of amyloid plaques in the brains of patients with AD (Connor et al., 1992) and its accumulation in brain is accompanied by increased oxidative stress which is also consistently observed in AD (Connor et al., 1992; Liu et al., 2005; Markesbery, 1997; Maynard et al., 2005). Iron is thought to induce its toxic effects by generating reactive oxygen species (ROS) via the Fenton reaction (Liu et al., 2006; Okada, 1996). The interaction of HFE with the cell surface-bound transferrin receptor (TfR) limits iron uptake into cells by reducing the affinity of the receptor for iron-bound transferrin (Tf). The H63D variant of HFE is unable to reduce affinity of the receptor for iron-bound Tf which results in uptake of more iron by cells. Indeed, MRI studies have revealed an increased accumulation of iron in carriers of H63D-HFE (Bartzokis et al.; Burt et al., 1998), and in individuals with hemochromatosis (Berg et al., 2000; Nielsen et al., 1995). More iron in cells can result in production of toxic oxidative radicals. Oxidative stress resulting from iron accumulation has been linked to synapse loss and cell death (Bush, 2003). These observations led to studies of proteins involved in iron homeostasis and their possible roles in AD with interest focused on HFE and its variants due to their prevalence, as high as 24%, in certain populations (Adams et al., 2005; Burt et al., 1998). A number of studies have reported an increased frequency of the H63D- HFE variant in AD patients (Combarros et al., 2003; Connor and Lee, 2006; Moalem et al., 2000; Percy et al., 2008; Pulliam et al., 2003; Sampietro et al., 2001). However, other studies showed a trend but failed to show significant association between H63D-HFE and AD (Berlin et al., 2004; Lleo et al., 2002). The apparently inconsistent findings may

92 reflect additional gene/environment interactions. Iron exposure can be affected by differences in life style (ex. smoking and diet) as well as other genetic factors. A consistent observation is that when the H63D-HFE variant is present in combination with the ApoE4 allele there is a significant increase in earlier onset of AD (Combarros et al., 2003; Percy et al., 2008; Pulliam et al., 2003).

Expression of ApoE4 in mice is associated with a lack of functional ApoE and a reduction in cholesterol delivery to neurons (Gong et al., 2002) which is consistent with findings in human AD patients (Lambert et al., 2005). Loss of cholesterol content found in AD brains (Ledesma, 2003; Ledesma and Dotti, 2006) can affect cell membrane integrity, formation of lipid rafts within cell membranes, and myelination. It can also result in synapse loss, altered cognitive function, and cell death. In light of the epidemiological studies showing earlier onset of AD in individuals expressing both APOE4 and H63D-HFE, we investigated the effect of expression of H63D-HFE on cholesterol metabolism and its possible association with development of the AD phenotype. More specifically, we tested the hypothesis that H63D-HFE is associated with the development of cognitive impairment by disrupting cholesterol metabolism thereby causing decreased brain volume, and increased memory deficits. To address this hypothesis, H63D-HFE expressing human neuroblastoma cells and a mouse model expressing H67D-HFE (analogous to the human H63D variant) were studied. In the present study, we found that the H63D/H67D-HFE variant induced reduced expression of cholesterol, altered expression of proteins involved in management of synaptic transmission and myelination, and poorer memory function.

Experimental Procedures:

Cell culture

Human neuroblastoma SH-SY5Y cells stably transfected to express either WT- or H63D- HFE were generated as previously described (Lee et al., 2007). Cells were cultured in DMEM-F12 medium (Gibco) supplemented with 10% fetal bovine serum (Gemini-

93 Bioproducts), 1% non-essential amino acids (Sigma), 1% penicillin-streptomycin (Invitrogen), sodium bicarbonate (Sigma), and Geneticin (Invitrogen). Cells were cultured at 37°C in an atmosphere of 5% CO2/95% air.

Gene expression analysis

Total RNA was isolated from SH-SY5Y cells expressing either WT- or H63D-HFE using the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Recovered RNA was quantified by measuring absorbance at 260 nm. Using the RT2 First Strand Kit (Qiagen), cDNA was prepared by reverse transcription of isolated RNA. For qRT-PCR reactions, 4 ng cDNA/ well was added to a commercially available 96 well Lipoprotein signaling and cholesterol metabolism gene array containing primer pairs specific for 96 genes involved in cholesterol metabolism (cat # PAHS-080Z; SABiosciences, Frederick, MD) along with RT2 SYBR® Green/ROX™ qPCR Master Mix and reactions carried out according to the manufacturer’s directions. Gene expression of each of the 96 genes was monitored using an ABI 7800 (Applied Biosystems, USA). Data were normalized to expression of mRNA for both actin and ribosomal protein L13A.

Biochemical assays

i) Cholesterol measurement

Lipids were extracted from equal numbers of cells (106) of each cell type using chloroform: isopropanol:Triton X-100 (7:11:0.1). The CHCl3 (bottom) layer was collected, dried, and the extracted material resuspended in the cholesterol reaction buffer provided with the Biovision kit. Cholesterol content was measured using the cholesterol quantification kit according to the manufacturer’s (Biovision) protocol.

ii) Determination of the effect of 24S-hydroxy cholesterol (24S-HC) or simvastatin on cell survival

Cells treated with 24S-HC (Enzo Lifesciences) or simvastatin (Sigma) were seeded at 105 cells per well in a 96 well plate and allowed to grow for 1 day in regular medium (as

94 described above). Cells were then treated with 50µM 24S-HC or 10µM simvastatin for 24 hrs in serum-free medium. Cell survival was measured using the MTS assay according to the manufacturer’s protocol (Promega).

iii) Western blot analyses

Cell lysates from WT- and H63D-HFE expressing cells were prepared in RIPA buffer (Sigma) containing a protease inhibitor cocktail (Sigma, S8340). Samples were sonicated for 10 seconds followed by centrifugation at 10,000g for 1 min at 4˚C and the supernatant collected. After determining protein content using the Dc Protein assay (BioRad), samples were diluted in sample buffer (137mM Tris-HCL, pH 6.8, 22% glycerol, 4.4% SDS, 5% bromophenol blue) to a final concentration of 50µg and boiled for 5 min prior to separation of protein components on a 10% pre-cast gel (BioRad) by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Mouse brain homogenates were prepared as described below and samples containing 30µg of total protein separated in the same manner. Following electrophoresis, proteins were transferred onto a PVDF membrane (Millipore) using a wet transfer apparatus (BioRad) with Tris-Glycine, pH 7.4, as the transfer buffer. After allowing the transfer to occur for ~16 hr at 4˚C, membranes were blocked with 5% non-fat dry milk in TBS-T buffer (1M Tris, 1.54M NaCl, pH 7.4, and 0.1% Tween-20) for 1 hr at room temperature prior to probing with primary antibodies overnight at 4˚C. Membranes were then washed three times with TBST prior to a one hr incubation at room temperature with the appropriate secondary antibody in TBS-T containing 2% non-fat dry milk. After washing membranes three times with TBST, bands were visualized using either Pico or Femto SuperSignal West Maximum Sensitivity SubstrateTM (Pierce) and exposure in a Fuji Film LAS 3000. Antibodies used were: CYP46A1 (ProteinTech; 12486-1-AP), HMGCoA reductase (Millipore; 07-457 and Biovision; 3952), Synaptophysin (ab8049), SNAP-25(ab53723), PSD-95 (ab18258), LDLR (ab30532), CNPase (ab6319), Caspr (ab34151), and NogoA (ab62024) (Abcam), Caspase-3 total and cleaved Caspase-3 (Cell signaling; 9662 and 9661), NeuN (Millipore; MAB377X), and Actin (Sigma; A5441). All primary antibodies were diluted 1:500 except actin (1:1000). Actin was used as a loading control. Horseradish peroxidase conjugated anti-mouse or anti-rabbit secondary antibodies

95 (1:5000, Amersham Bioscience) were used to monitor bands bound by primary antibodies. Comparisons were made between samples run on the same blot. Band intensity was measured using MultiGauge software (FujiFilm). Expression of each protein studied was normalized to actin and given as a percentage of the samples from WT-HFE cells/animals. Protein expression studies focused on mice 18 and 24 months of age because they were the mice used for behavioral studies and brain volume measurements.

Immunocytochemistry:

For a qualitative analysis of cellular cholesterol, cells were stained with filipin which is known to bind cholesterol (Miller, 1984). Briefly, 105 cells were seeded per well on collagen coated 4 well chamber slides and maintained as described previously. When just confluent, cells were washed with phosphate buffered saline (PBS) prior to fixation in 4% paraformaldehyde in PBS (30 min, 4˚C). After washing cells with PBS, cells were stained with filipin (50µg/ml) for 5 min in the dark. After removing excess filipin by washing the cells three times in PBS cells were visualized by fluorescent microscopy using a Nikon Eclipse 80i microscope connected to a Nikon Digital Sight camera. Images were captured using Nikon’s NIS-Elements Br Software,version 3.00.

Animals

All animal experiments were approved by the Penn State Hershey Medical Center IACUC (protocol #2007-131). H67D knock in mice were generated and genotyped as previously described (Liu et al., 2011; Tomatsu et al., 2003). For harvesting tissue, mice were anesthetized with pentobarbital (>100 mg/kg, i.p. administration) or a ketamine and xylazine mixture (>100 mg/kg ketamine, 10 mg/kg xylazine, i.p. administration). Blood was collected by cardiac puncture. Serum was collected by centrifuging samples (20 min at 12000rpm) at 4˚C and was stored at -800C until analyzed. At the end of behavioral testing, brain and other organs were collected. The right half of each brain was snap frozen in liquid nitrogen and stored at -80˚C until studied. The left halves of each brain were fixed overnight in ice cold 4% paraformaldehyde prior to embedding in paraffin and

96 then stored at room temperature until analyzed. Brains of both male and female mice of each genotype were examined at 12, 18, and 24 months of age. For each of the behavior tests, mice were randomized into groups of 3-4. Mice were kept on a 12 hr light/12 hr dark cycle.

Preparation of Mouse brain homogenates

Right brain hemispheres of mice were weighed and homogenized in 0.1M phosphate- buffered saline (PBS), pH7.4, containing 0.32M sucrose and a cocktail of protease inhibitors (Sigma), at 4˚C using a dounce homogenizer. For brain cholesterol analyses, brains were homogenized in the same buffer minus 0.32M sucrose.

Histological Analyses for determination of neurodegeneration

Serial 8µm-thick coronal brain sections of WT- and H67D-HFE homozygous mice (stratified n=3 per gender genotype group) between Bregma -2.30mm to -2.46mm were used for immunohistochemical analyses to assess apoptotic prevalence in the cortex and hippocampus.

i) Immunofluorescent staining of cleaved Caspase-3 and neuronal nuclei

Paraffin-embedded 8µm-thick brain sections of male and female mice were deparaffinized by heating at 55˚C for 25 min and subsequent immersion in xylenes. Tissue sections were rehydrated in a series of ethanol dilutions ending with distilled water. Sections were then boiled for five min in 10mM citrate buffer, pH 6.0, prior to blocking in an aqueous solution containing 10% normal goat serum (NGS; Gibco) and 0.1% Triton X-100 (Sigma) for one hr at 23˚C. Sections were then exposed to rabbit anti- cleaved caspase-3 polyclonal (Cell Signaling) and mouse anti-neuronal nuclei (NeuN) monoclonal (Chemicon) antibodies, diluted 1:300 and 1:100, respectively, in a solution containing 10% normal goat serum (NGS; Gibco) and 0.1% Triton X-100 (Sigma). After 18 hrs at 4˚C in a humidified chamber, unbound primary antibody was removed by 4,

97 10min washes with PBS, pH 7.4. Sections were then incubated for one hr at room temperature with goat anti-rabbit antibody conjugated to AlexaFluor 555 (Invitrogen), goat anti-mouse antibody conjugated to AlexaFluor 488 (Invitrogen), and DAPI (4',6- diamidino-2-phenylindole; Invitrogen) at respective dilutions of 1:200, 1:300 and 1:10,000. Unbound antibody was washed from sections using PBS as described above, a drop of Fluoromount (Sigma) applied per section and sections sealed with glass coverslips. Slides were stored in the dark at 4˚C and visualized within 48 hr. Specificity of both NeuN and cleaved caspase-3 antibodies was confirmed by Western blotting.

ii) TUNEL assay

The DeadEnd Fluorometric TUNEL System (Promega) was used according to the manufacturer’s protocol. Sections were deparaffinized via immersion in xylenes and rehydrated as described above. Sections were then fixed for 15 min in a 4% paraformaldehyde solution (Sigma) and washed in PBS. Tissues were then permeabilized by a 10 min incubation with 20 µg/ml Proteinase K. After re-fixing sections in 4% paraformaldehyde, they were washed with PBS. A positive control was prepared by incubating one slide for 10 min with 10 units/ml of DNase 1(Qiagen). Tissues were incubated in equilibration buffer, followed by a two hr incubation in the dark at 37˚C with recombinant terminal deoxynucleotidyl transferase and a fluorescein conjugated dUTP mix. The reaction was terminated by immersion in a neutral citrate buffer, and unincorporated fluorescent nucleotides removed by washing with PBS. Nuclei were stained by a 15 min exposure of sections to a 1µg/ml solution of propidium iodide. Sections were then washed thoroughly with distilled water prior to being sealed, and processed as described previously.

iii) Immunofluorescent imaging and analysis

Immunofluorescent brain sections were visualized using a Nikon Eclipse 80i microscope connected to a digital camera (Nikon Digital Sight). Images were captured using Nikon’s NIS-Elements Br Software, version 3.00 at 23˚C in a dark room. The cleaved caspase-3

98 and neuronal nuclei colocalization study provided qualitative information, while the TUNEL assay provided quantitative data.

The number and proportion of healthy versus apoptotic cells were quantified by a blinded investigator using the Image J Cell Counter to analyze a 431µm by 323 µm region of the posterior parietal cortex. Similarly, apoptotic cells were quantified in the hippocampal CA3c area between dentate gyrus pyramidal cell layers in a 150 µm by 325 µm region. Apoptotic cells were defined by the colocalization of TUNEL and propidium iodide staining which gave a yellow hue when the images were merged. Brightness and contrast were adjusted equivalently across all images to reduce background (Adobe Photoshop 7).

Brain volume measurement

Changes in brain volume were measured using 3D T2-weighted scan rapid acquisition with relaxation enhancement (RARE) magnetic resonance imaging (MRI). Mouse brain images were obtained using a 7-tesla (300MHz) 70/20as Bruker Biospec MRI system (Bruker Biospin, Germany). Prior to placement in the animal holder (an Acrylic Glass® animal cradle equipped with a nosecone and adjustable bite bar), mice were anesthetized using vaporized isoflurane gas (2-3%), and their eyes lubricated with a sterile ophthalmic solution. During imaging, animals were exposed to isoflurane gas (1- 2%, 0.8 liters/minute flow rate). To ensure the animal’s safety during the imaging procedure respiration was monitored using a PC-SAM Model 1025 fiber optic sensor (SA Instruments, Stony Brook, NY). The flow of the isoflurane gas was adjusted during imaging in response to the animal’s monitored breaths per minute. A vortex heating system (Vortex Tube, Exair Corporation, Cincinnati, OH) was used to warm the animals during imaging. Upon completion of the imaging, the animal was removed from the MRI system and allowed to recover underneath a 60-watt frosted light bulb. Baseline images were obtained when the mice were 9 months old. The same animals were scanned again at 22 months.

After the initial localizer scan was performed, a 3D T2-weighted scan (RARE) of the entire brain was acquired using the following parameters: TR/TE=2000/16.75ms, 2

99 averages, 4 echoes, 256x256x32 matrix, 2.5x2.5x1.5cm FOV, RARE factor=8 and slab thickness 15mm. Total acquisition time of the scan was 75 min. Images were processed by qMRI software followed by region of interest (ROI) analysis using MRICro. Quantitative analyses were performed to obtain total brain volume, ventricle size, and thickness of corpus callosum.

Learning and memory assessment tests i) The novel object recognition test (Kenney et al., 2011) was used to measure the ability of mice to remember the difference between an old and a new object. Mice with intact memory prefer to spend more time with a novel object than with an object they have seen before. The outcome measure for this test is the time spent with each object. The objects used in this study included a 50ml plastic conical tube, 25ml glass flask, square-shaped plastic blocks, and small plastic rectangular container. Briefly, on day 1, each mouse was habituated to a square open field (50cm by 50cm) for 7 min. The apparatus was cleaned with 70% ethanol between the testing of each subject. On day 2, each mouse was allowed to explore two identical objects for 7 min in the same square area. Now both the apparatus and objects were cleaned with 70% ethanol between the testing of each subject. An hour later mice were exposed to the same environment except that one of the old objects was replaced with a novel object. The amount of time spent with each object was measured over a 7 min period. This provided a measure of short- term memory. On day 3, 24 hr after the mice first saw the objects, object 3 (which was seen during the post-1hr test period) was replaced with a new novel object. Again, the amount of time spent with each object during a 7 min test period was recorded. This measure provided an assessment of long-term memory. Test sessions were recorded and analyzed using a video camera and Limelight behavior tracking analysis software (Actimetrics). Object exploration was counted when a mouse was facing towards the object and sniffing and/or standing near it (<2cm). Preference for the novel object was calculated by dividing time spent with the novel object by the sum of time spent with both the familiar and the novel object. Both male and female mice (n=7-17 per genotype) were used in the study.

100 ii) The novel object location test was performed in the same manner as the novel object recognition test except, rather than using a novel object, the location of the object was changed during the test phase. Different sets of objects were used for the object recognition and object location tests to ensure that the animals did not get too familiar with them. Visual cues consisting of a big checkered pattern and a triangular pattern were inserted on opposite walls of the open filed apparatus (Kenney et al., 2011). iii) The Barnes maze test was used to assess spatial learning and memory (Patil et al., 2009). The maze consisted of a circular table with 20 holes, with an escape box, placed under one hole (Stoelting Co.). When placed on the maze, mice were motivated to search for the hole leading to the escape box via exposure to a bright light (80-watts bulb; 90cm above maze) and a loud buzzer (85dB). Mice were pre-trained with use of an escape box prior to testing. The position of the escape box was kept constant throughout the testing. To prevent a mouse from following smell cues from a previous mouse, the surface of the Barnes maze, start chamber, and escape box were washed down with 70% ethanol between each mouse. Each mouse went through 4 trials per day for 4 days with a 15 min inter-trial interval. During each trial, the mouse was placed in the center of the maze under a start chamber for 10 seconds with both the buzzer and light on, then allowed to explore the maze for 3 minutes. A probe trial on day 5 was used to measure short term memory, while a probe trial on day 12 measured long term memory. For both probe trials, mice were allowed to explore the maze for 90 seconds. All sessions were video recorded. In all trials, time for the mouse to find the escape hole, number of errors, and distance traveled to reach the escape hole were measured. Primary latency was defined as the time it took for the mouse to first visit the escape hole. Total latency was the time it took for a mouse to enter the escape hole. Latency measures were scored manually. Other outcome measures were analyzed using ANY-maze (Stoelting Co.) video tracking software.

Statistical analyses

Data were analyzed using a one-way analysis of variance (ANOVA) followed by Newman-Keuls post-hoc tests to evaluate effects of multiple treatments. A two-way

101 ANOVA followed by Bonferroni tests was used to compare the effect of age and genotype for age-related changes in brain cholesterol of HFE variant mice. Two-tailed and unpaired Student t-tests were used to compare the effect of genotype using Graphpad (Prism) software. Behavioral data were analyzed using mixed factorial ANOVAs using Statistica. Post hoc tests were conducted, where appropriate, using Newman-Keuls tests. Correlation analyses were performed using Graphpad prism. Error bars in all figures represent the standard error of the mean. Symbols used to indicate significance were as follows; *** p<0.001, ** p<0.01, * p<0.05.

Brain iron measurement

Total brain iron content was measured by graphite furnace atomic absorption spectroscopy (spectrophotometer model 5100 AA, Perkin-Elmer, Norwalk, CT) according to standard methods (Erikson et al., 1997). Protein content was measured by Lowry assay according to the manufacturer’s protocol (Bio-Rad) / BCA assay (Pierce). Iron content was calculated per µg of protein.

RESULTS

Effect of the H63D-HFE variant on total cholesterol content

To study the effects of H63D-HFE on cells, SH-SY5Y cells transfected to stably express WT- or the H63D-HFE protein (Lee et al., 2007) were used. Because previous studies indicated that both proteins are associated with lipid rafts and that expression of H63D- HFE was accompanied by changes in expression of glycosphingolipids (Ali-Rahmani et al., 2011), lipids enriched in lipid rafts, we hypothesized that it might also affect metabolism of cholesterol, another lipid enriched in rafts. Analysis of cellular cholesterol content indicated that H63D-HFE expressing cells had ~50% less cholesterol than WT- HFE cells (Figure 1A). Exposure to filipin, a cholesterol binding dye (Miller, 1984), also indicated that H63D-HFE cells had less cholesterol than those expressing WT-HFE (Figure 1B).

102 Effect of H63D-HFE on expression of proteins affecting cholesterol metabolism

To investigate the underlying cause(s) of the decrease in cholesterol content of H63D- HFE cells relative to those expressing WT-HFE, expression of 96 key genes encoding proteins involved in cholesterol metabolism was ascertained using a Lipoprotein signaling pathway focused gene array. Expression of a number of genes was altered in H63D-HFE cells relative to WT-HFE cells (Table 1). Only genes with a 2-fold or greater change in expression were considered to be significantly affected. CYP46A1 and AMPK, the two most highly over-expressed genes (>5 fold), encode the enzymes cholesterol 24- hydroxylase and AMP-activated protein kinase, respectively.

Protein expression of CYP46A1 was also elevated in H63D-HFE cells (Figure 1C). Because activation of AMPK results in inhibition of cholesterol synthesis (Hardie, 2003), we measured and found that the expression of HMGCoA reductase (HMGCoAR), an enzyme mediating the rate limiting step in cholesterol synthesis, was reduced in H63D- HFE cells (Figure 1D). Though the level of mRNA for HMGCoAR wasn’t significantly altered in the gene expression study, protein levels were significantly lower in H63D- HFE cells. No difference in expression of the low density lipoprotein receptor (LDLR) was observed between WT- and H63D-HFE expressing cells, indicating that the uptake of cholesterol by cells was not altered. To determine whether there was an effect of iron on the cholesterol content of cells, WT- and H63D-HFE cells were treated with 100µM of the iron-loading compound, ferric ammonium citrate (FAC), or 20µM of the iron chelator, deferoxamine (DFO) for 24 hr and cholesterol content measured (Figure 1F). To demonstrate that cells took up iron when treated with FAC and were depleted of iron upon treatment with DFO, cellular expression of TfR was measured under the same treatment conditions used to assess cellular cholesterol content (Figure 1E). The expression pattern of TfR for WT- and H63D-HFE cells after treatment with FAC or DFO indicated that both treatments were effective; iron-loaded cells showed a significant reduction in expression of TfR, while iron depleted cells had a significant increase in its expression relative to untreated controls (p<0.001). Genotype-specific effects were observed on cholesterol content of cells after manipulation of their iron content. While iron-loading was accompanied by an increase in cholesterol content in WT-HFE cells (p

103 <0.05), H63D-HFE cells underwent a significant reduction in cholesterol content (p <0.001) with iron loading treatment relative to the untreated controls (Figure 1F). While iron chelation significantly elevated cholesterol of WT-HFE cells (p <0.05), cholesterol content of H63D-HFE cells was not significantly altered relative to untreated controls (p >0.05). These observations indicate that a decrease in iron in H63D-HFE cells was accompanied by an elevation in cholesterol content relative to that of iron-loaded H63D- HFE cells (p <0.001), suggesting that in these cells there was an inverse effect of iron content on cholesterol level.

Effect of cholesterol manipulations on survival of H63D-HFE cells

To ascertain whether reduction of cholesterol in H63D-HFE cells had a functional consequence, survival of cells treated with simvastatin (an inhibitor of HMGCoAR and therefore of cholesterol biosynthesis) was measured using the MTS assay. Decreased survival was observed in H63D-HFE cells treated with simvastatin relative to WT-HFE controls (Figure 2A), indicating that greater reduction of cholesterol in H63D-HFE cells decreased their survival. To determine whether 24S-hydroxycholetsterol (24S-HC), the product of the enzymatic action of CYP46A1, affected cell survival, cells were treated with 50µM 24S-HC in medium containing no serum and cell survival also measured using the MTS assay. Survival of H63D-HFE cells treated with 24S-HC was reduced by 10% compared to WT-HFE cells treated analogously and by 30% relative to untreated control WT- and H63D-HFE cells, indicating genotype dependent growth inhibiting/cytotoxic effects of 24S-HC (Figure 2B).

Effect of H67D-HFE (mouse homologue of H63D-HFE) on cholesterol metabolism in brains of mice

To test the relevance of the in vitro findings, total cholesterol levels in the brains of transgenic mice expressing either WT- or H67D- (homologous to human H63D) HFE were determined for mice 12, 18, and 24 months of age. With increasing age, a

104 significant reduction (p<0.0001, n=7-19, two-way ANOVA followed by Bonferroni post- tests) of total cholesterol was observed in brains from H67D-HFE mice relative to those from WT-HFE mice (Figure 3A). Consistent with findings from the in vitro studies, protein expression of CYP46A1 was elevated (p<0.05) while HMGCoAR expression was significantly (p <0.001) reduced in total brain homogenates of 18 and 24 month old H67D-HFE mice (Figure 3B & C). No significant difference in expression of LDLR was observed (Figure 3D). Due to the similar trend of changes in protein expression, only graphs from 18 month old animals are shown. Combined these changes are consistent with the decrease in total brain cholesterol found in H67D-HFE mice.

Effect of H67D-HFE on expression of synaptic proteins

Because cholesterol plays a critical role in synaptic vesicle recycling, regulation of synaptogenesis, and neurite outgrowth (Koudinov and Koudinova, 2005; Linetti et al., 2010) and the association of cholesterol with synaptophysin is essential during synaptogenesis (Thiele et al., 2000), we hypothesized that a decrease in cholesterol would disrupt synaptic machinery in H67D-HFE mice. To determine whether the H67D-HFE induced reduction in brain cholesterol had an effect on synaptic transmission, we measured total brain expression of pre- and post-synaptic proteins. Western blot analyses indicated that brains from H67D-HFE mice had reduced levels of the presynaptic vesicle proteins synaptophysin (p<0.001) and SNAP-25 (p<0.01), and post-synaptic vesicle protein PSD-95 (p<0.01) relative to those from WT-HFE mice (Figure 4A-C). Total brain expression of SNAP-25 positively correlated with brain cholesterol content (overall: Pearson r=0.7263, p<0.001 and a separate analysis for H67D-HFE mice alone: Pearson r=-0.8578, p<0.01) (Fig. 10A).

H67D-HFE affects myelination

Because cholesterol is essential for proper myelination of the brain (Saher et al., 2005), we monitored expression of NogoA, Caspr and CNPase, markers of myelination, in

105 homogenates of brains from 18 month old mice. While expression of NogoA and Caspr was significantly decreased in brains from H67D-HFE mice (p<0.001 and 0.01, respectively), CNPase expression was not (p>0.05) (Figure 5A-C). Interestingly, volume of the corpus callosum, a region of the brain primarily involved in myelination (Yates and Juraska, 2007), was significantly (p<0.001) reduced in brains of 24 month old H67D- HFE mice (Figure 7A & B).

Expression of H67D-HFE is accompanied by neuronal loss and a reduction in brain volume

The significant reduction in expression of both synaptic and myelin proteins supported the hypothesis of increased neurodegeneration in brains of H67D-HFE mice. Reduction in expression of the neuronal marker NeuN in brains of H67D-HFE mice (p<0.001) coupled with the concomitant increase in levels of total and cleaved Caspase-3 in whole brain homogenates (Figure 6 A-C) further supported this hypothesis. To ascertain whether the reduction in neuronal markers was due to neuronal loss, hippocampal sections from brains of 18 month old mice were analyzed for NeuN (nuclear neuronal protein) and cleaved Caspase-3 using immunohistochemistry. The colocalization of NeuN and cleaved Caspase-3 indicated neuronal loss in the cortex and hippocampus of H67D-HFE mice (Fig. 6D & E), a loss corroborated by the increase in TUNEL staining (Fig. 6F & G) seen in the cortex and hippocampus of brains from 18-month old H67D- HFE mice. Quantitative analyses showed ~90% increase in TUNEL positive apoptotic cells in the cortex and 31% in the hippocampus of H67D-HFE mice relative to WT-HFE mice (p= 0.034 and p= 0.015, respectively) (Figure 6H).

Because excessive neuronal loss can result in brain atrophy, MRI was used to measure brain volume of WT- and H67D-HFE mice. Baseline images obtained from 9 month old mice as well as follow-up scans taken at 22 months, showed 6 and 6.5% decrease in total brain volume of H67D-HFE mice relative to brains of WT-HFE mice (Fig. 7B; p= 0.02 and 0.002; unpaired t-test used at 9 (n=10 per genotype) and 22 months (n=8-9 brains per genotype), respectively. Region of interest (ROI) analyses in the brains

106 of 22 month old mice showed a 57% increase in ventricle size (p= 0.006) and 22% decrease in volume of the medial corpus callosum (p=0.0007) in the brains of H67D-HFE mice (Figure 7 A&B). Consistent with volumetric measurements, total weight of brains from H67D-HFE mice was significantly less than brains from WT-HFE mice (Fig. 7C; p<0.0001). Overall, brain volume measured by MRI and wet brain weight correlated positively (Pearson r=0.53; p=0.03) (Fig. 10B). Brain cholesterol content also correlated positively with brain volume overall and by genotype (overall: Pearson r=0.91; p<0.001, WT-HFE: Pearson r=0.86; p=0.0024, H67D-HFE: Pearson r=0.81; p=0.02, respectively) (Fig. 10C).

H67D mice show impaired recognition and spatial memory

Novel object recognition, novel object location, and the Barnes maze were used to assess functional effects of the H67D-HFE variant on learning and memory. These tests were performed in mice at 14 or 18 months and, in a separate group of mice, at 24 months. Both male and female mice were used for all three tests and data were collapsed across male and female mice to show the effects of genotype (Figure 8). Effects of gender, where found, are discussed below. For the novel object recognition and location tests, testing occurred 1 and 24 hr following initial exposure to the objects and location. Data were collapsed across male and female mice and across the 1 and 24 hr tests because performance of both genders in these tests trended in the same direction (Figure 8A & B); exceptions are discussed below.

Novel Object Recognition: 14 Months. The data were analyzed using a 2 x 2 x 2 mixed factorial ANOVA varying gender, genotype, and days (1 or 24 hr test). The results of this ANOVA revealed a significant main effect of genotype, F (1,15) = 10.64, p<0.005, showing that H67D-HFE mice exhibited less preference for the novel object (i.e., poorer retention) compared to WT-HFE mice overall. No other main effects or interaction was statistically significant; p<0.05 (Figure 8A, left panel). Novel Object Recognition: 24 Months. A similar pattern was evident at 24 months, (Figure 8A, right panel). In support,

107 the results of a 2 x 2 x 2 mixed factorial ANOVA showed that the main effect of genotype was significant, F (1,13) = 19.7, p<0.0006. Thus, like the 14-month old animals, 24 month old H67D-HFE mice exhibited impaired recognition of a novel object. Unlike the 14 month old subjects, a significant genotype x gender interaction was observed in 24 month old mice, F (1,13) = 5.74, p<0.03. Post hoc Newman-Keuls tests indicated that females in the H67D-HFE group actually performed more poorly than the males, p<0.02 (data not shown). Indeed, female H67D-HFE mice performed significantly more poorly than did their WT-HFE controls, p<0.001, while male mice in the H67D- HFE group did not (p>0.05). Therefore, the effect in the 24-month old subjects was carried primarily by the female H67D-HFE mice.

Preference for a Novel Location: 14 months. Data for these studies were analyzed as described for the novel object recognition test. The results showed no statistically significant main effect of gender, F (1,17) = 2.21, p<0.1, genotype, F (1, 17) = 3.27, p<0.09, or days, F < 1.0, nor interactions thereof, Fs < 1.0, (Figure 8B, left panel). Preference for a Novel Location: 24 months. The results of a similar ANOVA for the 24-month old mice revealed a highly significant main effect of genotype, F(1,14) = 14.1, p<0.002, indicating that the aged H67D-HFE mice had poorer recollection of familiar objects when presented in a novel location than WT-HFE controls (Figure 8B, right panel). Neither the main effect of gender, F (1,14) = 2.84, p<0.1, nor the gender x genotype interaction, F < 1.0, was significant. In contrast, the main effect of day, F (1,14) = 11.22, p<0.004, and the day x genotype interaction, F (1,14) = 4.79, p<0.05, were statistically significant. Post hoc analysis of the day x genotype interaction showed that WT-HFE mice performed better than H67D-HFE mice, but only when tested 1h following initial exposure to the objects, p<0.05 (data not shown). No other effects were significant.

Barnes Maze Primary Latency: To assess spatial memory, responses of 18- and 24- month old mice were assessed using a non-aversive Barnes maze test. Analysis of data obtained on days 1-4 was conducted using a 2 x 2 x 4 mixed factorial ANOVA, varying genotype, gender, and day (1-4). Primary Latency: 18-month old mice. The results

108 revealed a significant main effect of genotype, F(1,28)= 15.93, p=0.0004, indicating that the H67D-HFE mice took longer to find the hole overall relative to WT-HFE controls ( Figure 8C). Neither the main effect of gender, F (1,28) = 2.84, p<0.1, nor gender x genotype interaction, F < 1.0, were significant. While the main effect of days, F (3,84) = 56.7, p<0.0001 was significant, the days x genotype interaction, F < 1 was not. Primary Latency: 24 month old mice. Analysis of the data for the 24 month old mice also revealed a significant main effect of genotype, F(1,17)= 19.31, p=0.0004 ( Figure 8F), and no significant main effect of gender, F<1. In contrast to 18-month old animals, the genotype x gender interaction F(1,17) = 4.69, p<0.05 was significant. Post hoc analysis of the genotype x gender interaction indicated that male H67D-HFE mice took significantly longer to find the hole than male WT-HFE mice p=0.001. The performance of female mice however did not differ as a function of genotype p>0.05. Together these observations indicate that the overall difference in primary latency was driven, primarily, by behavior of the male mice. The main effect of days, F (3,51) = 17.93, p<0.0001, but not the days x genotype interaction, F<1.0 attained statistical significance.

Barnes Maze Total Latency: Total latency was defined as the amount of time it took for the mouse to enter the target hole. Total Latency: 18 month old mice. Analysis of data as described for tests of primary latency indicated a significant main effect of genotype, F(1,28)= 5.23, p<0.03 in 18 month-old mice, reflecting the fact that the H67D-HFE mice took longer to enter the hole than WT-HFE controls ( Figure 8D). Neither the main effect of gender, F(1,28)=3.26, p<0.1, nor the gender x genotype interaction, F < 1.0, was significant. The main effect of days, F(3,84)=29.7, p<0.0001 and the days x genotype interaction, F(3,84)=2.82, p=0.04 were significant (data not shown). Post hoc analysis of the day x genotype interaction showed that the differences in total latency between WT- and H67D-HFE mice were not statistically significant on a day by day basis, but that there was an overall effect of genotype on performance across days. Interestingly, the total latency of WT-HFE mice began to improve on day 3 when compared with day 1, p<0.0004, while H67D-HFE mice did not exhibit improvement in total latency until day 4 relative to day 1, p < 0.05.

109 Barnes Maze Total Latency: 24 month old mice. A similar analysis of data obtained for 24-month old mice also indicated a significant main effect of genotype, F(1,17)= 10.36, p = 0.005 ( Figure 8G), but not of the main effect of gender, F (1,17) = 1.63, p=0.2, nor the gender x genotype interaction, F (1,17) = 1.97, p<0.2. As occurred with the 18-month old mice, the main effect of days, F (3,51) = 17.73, p<0.001, and the days x genotype interaction, F (3,51) = 7.37, p=0.0003, were statistically significant showing that the latency to get to the hole changed over trials as a function of genotype. Post hoc analysis of the day x genotype interaction revealed that the total latency of H67D-HFE mice was significantly higher than that of WT-HFE mice on day 4, p=0.004. As found for the 18-month old WT-HFE mice, total latency for the WT-HFE mice began to improve on day 3, relative to day 1, and was even shorter on day 4, p=0.002 and p=0.0001, respectively. Interestingly, the 24-month old H67D-HFE mice did not exhibit improvement in total latency for any day during the acquisition trials. These data indicate that at 24 months, H67D-HFE mice exhibit a significant decline in spatial learning ability and memory retention.

Barnes Maze Mean Latency for probe trials on days 5 & 12: Memory retention was assessed in probe trials and the data obtained on days 5 and 12 analyzed separately for 18 and 24 month old mice using 2 x 2 x 2 mixed factorial ANOVAs varying genotype, gender, and days. The results showed that, for both 18- and 24-month old mice, there was a significant main effect of genotype, F(1,28)=11.43, p=0.002 ( Figure 8E), and F(1,17)=16.81, p=0.007 ( Figure 8H), respectively. These results indicate that, at both ages, the H67D-HFE mice took longer to find the hole than WT-HFE controls. No other interactions were significant.

Barnes Maze Total Distance traveled: The data were analyzed as described for studies of primary latency using the Barnes maze. The total distance traveled by a mouse to find the target hole was measured using AnyMaze animal tracking software. For both 18- and 24-month old mice there was a significant main effect of genotype, F(1,30)=5.02, p=0.03,

110 and F(1,18)=12.72, respectively, indicating that H67D-HFE mice not only took longer to find the hole, they also traveled longer distances to find and enter the hole overall relative to WT-HFE controls ( Figure 9A & E). That said it is important to know that while differences in latency persisted on day 4 for example i.e. latency for H67D-HFE mice, the distance traveled between WT- and H67D mice did not differ on day 4, indicating that difference in latency is not due to mere difference in activity per se. In contrast to 18- month old mice for which neither the main effect of gender, F (1,30) = 2.0, p<0.2, nor the gender x genotype interaction, F(1,30)= 2.1, p<0.2 was significant, both the main effect of gender, F (1,18) = 8.13, p<0.01, and the gender x genotype interaction, F (1,18)=29.8, p<0.001, were significant at 24-months. Post hoc analysis of the genotype x gender interaction showed that 24-month old male H67D-HFE mice traveled a longer distance to find the target hole than WT-HFE male or female mice, p=0.0002. While the main effect of days was significant for both 18-month old, F (3,90) = 69.5, p<0.0001 and 24-month old F (3,54) = 13.42, p<0.0001, mice, the days x genotype interaction, F (3,90)= 1.5, p<0.2, 18 months; F (3,54)= 1.6, p=0.1, 24-months, was not. Interestingly, there was also a significant 3-way days x genotype x gender interaction, F(3,54)= 5.98, p=0.001 for 24- month old mice and post hoc analysis indicated that the male H67D-HFE mice traveled significantly more distance on days 1 and 3 p <0.05, while female H67D-HFE mice traveled a shorter distance than female WT-HFE controls.

Barnes Maze Mean distance traveled for probe trials on days 5 & 12: The mean distance traveled to reach the target hole was also assessed for probe trials and the data were analyzed separately for the 18 and 24 month old mice using separate 2 x 2 x 2 mixed factorial ANOVA varying genotype, gender, and days (5 and 12). For both 18- and 24-month old mice, there was a significant main effect of genotype, F(1,28)= 11.43, p=0.002 and F(1,28)= 17.07, p=0.0006, respectively, showing that overall the H67D mice took longer to remember and find the hole relative to WT controls ( Figure 9B & F). No other main effects or interactions were significant.

Barnes Maze Total Errors: 18 months. The data were analyzed using a 2 x 2 x 4 mixed factorial ANOVA varying genotype, gender, and acquisition trials on days 1-4. When an

111 animal poked his/her nose in a hole other than the target hole that was considered an error. For 18 month old mice, there was no significant effect of genotype ( Figure 9C), gender, or genotype x gender interaction F<1. However, the main effect of days, F (3,90) = 41.1, p<0.001 was significant but the days x genotype interaction F <1 was not. Barnes Maze Total Errors: 24 months. On the other hand, for 24 month old mice there was a significant main effect of genotype, F(1,18)= 8.08, p=0.01, showing that the H67D-HFE mice made more errors during acquisition trials overall relative to WT-HFE controls ( Figure 9G). Neither the main effect of gender, F<1, nor the gender x genotype interaction, F < 1.0, was significant. The main effect of days, F (3,54)=11.32, p< 0.001 was significant but the days x genotype interaction, F=1, did not attain statistical significance.

Barnes Maze Mean Errors during probe trials on days 5 & 12: For probe trials, the data were analyzed as described for analysis of Barnes maze mean distance data. Errors: 18 month old mice. No significant differences were observed in these variables or any interactions thereof, F<1 for 18-month old mice (Figure 9D). Errors: 24 month old mice. By 24 months, a significant main effect of genotype, F(1,18)= 9.84, p=0.005, was observed (Figure 9H), indicating that, overall, the H67D-HFE mice made more errors in finding the target hole than did the WT-HFE controls. While the main effect of days, F (1,18)=14, p=0.01 was also significant, no other interactions were, p> 0.05.

Collectively, data for all the measures tested here such as primary latency, total latency, mean latency, distance traveled and total errors made during acquisition and probe trials indicated that H67D-HFE mice had impaired spatial learning and memory and that some of the effects were more significant in older than younger mice. In addition, these data indicate that while female H67D-HFE mice did worse on recognition tasks, male H67D did worse on spatial tasks.

112 Correlations:

Correlation analyses were conducted to determine whether changes in cholesterol concentration, protein expression, or brain volume of 24 month old mice correlated with changes in their behavior. The results indicated that cholesterol content correlated positively with results obtained in the novel object recognition test (overall (both WT- and H67D-HFE): Pearson r=0.52, p=0.03; H67D-HFE: Pearson r=0.72, p=0.04) (Fig. 10F) and also for the novel object location test (overall: Pearson r=0.76, p=0.0002; H67D-HFE: Pearson r=0.76, p=0.02) (Fig. 10G). Overall, brain cholesterol content also correlated negatively with performance on the Barnes maze task. Lower cholesterol content was associated with higher primary latency (overall: Pearson r=-0.59, p=0.0005), higher total latency (overall: Pearson r= -0.4677, p=0.0325), and higher mean latency (overall: Pearson r= -0.81, p=0.0005), resulting in poorer performance (Fig. 10H-J).

Brain volume also correlated positively with performance on the novel object recognition task (overall: Pearson r=0.81, p=0.002 and for H67D-HFE: mice Pearson r=0.81, p=0.03) (Fig. 10K) and negatively with average primary latency on Barnes maze (overall: Pearson r=-0.74, p=0.006 and for H67D-HFE mice: Pearson r=-0.90, p=0.01) (Fig. 10L). Together, both of these indicate that smaller total brain volume is associated with impaired memory function.

Synaptophysin expression correlated positively with both the novel object recognition (overall: Pearson r=0.79, p=0.0007) (Fig. 10M) and novel object location tasks (overall: Pearson r=0.6569, p=0.0078) (Fig. 10N) and negatively correlated with primary, total and mean latencies on the Barnes maze task (overall: Pearson r=-0.63, p=0.01) (Fig. 10). This means that lower the expression of synaptophysin, lower preference for the novel object or novel location, and more memory impairment. Because higher latency on a Barnes maze task means impaired memory, the mouse takes longer to remember and find the target hole, a negative correlation of synaptophysin expression with latencies for example would mean that lower synaptophysin expression is associated with longer latency and more memory impairment. Results of these correlations indicate that lower total brain expression of synaptophysin is associated with impaired spatial (increased latency) and recognition memory (lower preference for the novel object).

113 SNAP-25 expression also correlated positively with novel object recognition (overall: Pearson r=0.93, p=0.0001 and for H67D-HFE: Pearson r=0.87, p=0.01) (Fig. 10O) and with novel object location tasks (overall: Pearson r=0.86, p=0.0001) (Fig. 10P).

Discussion:

In this study, we found that expression of H63D-HFE in cells and the mouse homologue, H67D-HFE, in mice, was accompanied by a significant reduction in both cell and brain cholesterol. The direct effects of the H63D-HFE variant on brain cholesterol metabolism are unknown but in our study the presence of this polymorphism was associated with a reduction in expression of synaptic proteins and proteins needed for myelination, an increase in neurodegeneration, and impaired learning and memory. This is the first study to show that a variant of HFE affects cholesterol metabolism and links the H67D-HFE phenotype to decreases in learning and memory. Tan et al. found that HFE knockout mice showed perturbed expression of genes involved in hepatic lipid metabolism (Tan et al., 2011). We have previously shown alterations in the sphingolipid metabolism of HFE variant cells (Ali-Rahmani et al., 2011). Both of these observations support our finding that expression of a variant of HFE (H63D) can affect cholesterol metabolism.

Epidemiological studies have shown that individuals with elevated plasma cholesterol may have an increased susceptibility to AD (Jarvik et al., 1995; Notkola et al., 1998; Roher et al., 1999). On the other hand, lower brain cholesterol levels have been seen in individuals who died with AD (Ledesma, 2003; Ledesma and Dotti, 2006). These apparent contradictions may reflect the fact that plasma cholesterol is not transported across the blood brain barrier, the brain must synthesize its own cholesterol (Dietschy and Turley, 2001). The increased probability of early onset of AD in people with both the ApoE4 allele and the H63D-HFE variant (Combarros et al., 2003; Percy et al., 2008) supports the hypothesis that dysregulation of brain iron and cholesterol may influence the etiology of AD. Excessive loss of brain cholesterol can be particularly devastating because the rate of brain cholesterol synthesis decreases with age under normal

114 conditions (Thelen et al., 2006a). Therefore, appropriate regulation of cholesterol metabolism is essential to prevent excessive cholesterol loss from the brain and the accompanying neurodegeneration. Because cholesterol is an integral component of lipid rafts, alterations in lipid rafts resulting from loss of cholesterol have been hypothesized to contribute to the loss of neural function and potentially their cell death in AD (Gellermann et al., 2005). Consistent with these studies, we found an age-related decrease in brain cholesterol content, and this decline was significantly greater in H67D-HFE mice than WT-HFE mice (Figure 3A). There was also evidence of increased neurodegeneration in the cortex and hippocampus of brains of H67D-HFE mice than those of WT-HFE mice (Figure 6). Both of these findings are in line with the loss of brain volume seen in the brains of H67D-HFE mice, demonstrating a link between HFE genotype, cholesterol, and neuronal survival.

Effect of H63D-HFE on genes involved in cholesterol metabolism

Using focused gene array analysis we found that CYP46A1 and AMPK were overexpressed in H63D-HFE expressing cells. These genes encode the enzymes cholesterol 24S-hydroxylase and AMP-activated protein kinase, respectively. Both gene products have been implicated in AD (Fang and Jia, 2007; Lopez-Lopez et al., 2007). CYP46A1 is a brain specific enzyme that catalyzes the conversion of cholesterol into 24S-hydroxycholesterol (24S-HC) and its expression can increase as a result of oxidative stress and traumatic brain injury (Lukiw, 2006). CYP46A1 mediated cholesterol hydroxylation is the only known mechanism whereby excess cholesterol can be cleared from the brain (Lutjohann, 2006). Elevation of 24S-HC has been reported in both serum and cerebrospinal fluid (CSF) from early stage AD patients (Leoni, 2009; Lutjohann et al., 2000; Papassotiropoulos et al., 2002). Higher protein expression of CYP46A1 correlates with a higher concentration of 24S-HC in the brain (Liao et al., 2010). In our study, the increase in CYP46A1 protein expression in H63D-HFE neuroblastoma cells and in the brains of H67D-HFE mice was accompanied by a reduction in expression of HMGCoAR, shown by others to be inhibited by 24S-HC (Cartagena et al., 2010). Combined these observations indicate that the decrease in cholesterol seen in H63D/H67D-HFE cells/mice correlated with increased expression of CYP46A1 and its

115 inhibition of HMGCoAR. The neurotoxic effects of 24S-HC, as reported by others (Yamanaka et al., 2011), were also recapitulated by our observation of decreased survival of H63D-HFE cells treated with 24S-HC. Another effect of elevated CYP46A1 is the disruption of glutamate transport caused by dissociation of EAAT2 from lipid rafts and the concomitant disruption of its function (Tian et al., 2010). We have previously shown altered glutamate transport in H63D-HFE cells (Mitchell et al., 2010).

Another gene AMPK that is overexpressed by H63D-HFE cells could also affect cholesterol metabolism. AMPK is a sensor of cell energy status and when energy stores are depleted, AMPK inhibits ATP-consuming processes such as cholesterol synthesis (Hardie, 2003) by inhibiting HMGCoAR. Overexpression of both, CYP46A1 and AMPK is consistent with our finding of reduced cholesterol levels in both H63D-HFE expressing cells and H67D-HFE expressing mice. The increased expression of AMPK could result from oxidative stress (Wu and Wei, 2012) and an increase in oxidative stress has been reported in both H63D-HFE cells (Lee et al., 2007; Mitchell et al., 2010), and in the brains of aging H67D-HFE mice (Wint et al. unpublished data) Collectively, these results indicate that expression of H63D/H67D-HFE induces an overall reduction in cholesterol content by inducing cholesterol efflux and reducing cholesterol synthesis.

Two mRNAs are down regulated in H63D-HFE cells consistent with reduction of cholesterol synthesis and transport. CYP7A1 encodes cholesterol 7 alpha-hydroxylase or cytochrome P450 7A1 (CYP7A1) and catalyzes the rate-limiting step in the synthesis of bile from cholesterol in liver. It is also thought to play a role in cholesterol elimination. Under conditions of low cholesterol, the expression of CYP7A1 is downregulated by the action of sterol regulatory element-binding proteins (SREBPs) (Kanayama et al., 2007). Another gene downregulated in H63D-HFE cells was ApoA1, which encodes apolipoprotein A1. ApoA1 is a major component of HDL and is involved in cholesterol transport (Millan et al., 2009). Kawano et al. (1995) found lower plasma levels of ApoA1 and ApoA2 in Japanese patients with late-onset non-familial AD. In addition, lower plasma ApoA1 levels and cognitive score have been reported in individuals with mild cognitive impairment. ApoA1 expression in plasma was also found to be the most

116 significant predictor of cognitive decline in elderly non-demented individuals (Song et al., 2012).

Impact of loss of cholesterol due to H63D-HFE

Reduction in the cholesterol content of H63D-HFE cells has multiple effects. Cholesterol is needed for cell growth and proliferation and we previously found that H63D-HFE cells grew more slowly than WT-HFE cells (Ali-Rahmani et al., 2011; Lee et al., 2007; Liu et al., 2011). Here we have shown that depletion of cellular cholesterol by treatment with simvastatin was particularly detrimental for the survival of H63D-HFE cells. Because statins are commonly prescribed for middle- and old-age individuals, a statin capable of crossing the blood brain barrier could be counter-indicated in individuals expressing the H63D-HFE variant. In addition to lowering brain cholesterol levels (Butterfield et al.; Kirsch et al., 2003; Lutjohann et al., 2004; Thelen et al., 2006b), statin treatment has been shown to increase the production of pro-inflammatory cytokines by microglia (Van Der Putten et al., 2011) and to impair re-myelination (Miron et al., 2009). Myelin status and cognitive performance have recently been linked in human studies (Bartzokis et al., 2006; Bartzokis et al., 2007). Consistent with these findings, we found disruption in the expression of myelin management proteins and memory impairment in our studies of H67D-HFE mice. Though further in vivo studies are needed to ascertain the detrimental effects of statin treatment in H67D-HFE variant mice, our findings from cell culture studies argue for assessment of these variables in human subjects.

Cholesterol is also an essential component of lipid rafts, membrane microdomains enriched in cholesterol, sphingolipids and proteins involved in signal transduction. Evidence for the need for membrane cholesterol is provided by the observation that membranes of cells from AD brains are less stable due to altered lipid composition (Ginsberg et al., 1993a; Ginsberg et al., 1993b). Alterations in the lipid components of lipid rafts have been hypothesized to affect cell signaling (Sottocornola et al., 2006) and protein-protein interactions thereby contributing to the loss of neural function and potentially the cell death seen in AD (Gellermann et al., 2005). We and others have

117 shown that HFE is found in lipid rafts. Though H63D-HFE is also associated with lipid rafts (Ali-Rahmani et al., 2011), it is possible that its interaction with other proteins and lipids may be affected due to the reduction in cholesterol content.

Cholesterol is essential for synaptic transmission (Linetti et al., 2010). Lack of cellular cholesterol can affect protein machinery responsible for regulating synaptic vesicle transport (Thiele et al., 2000) and can alter neuronal function. Consistent with that finding, we observed a reduction in expression of the raft-associated synaptic proteins synaptophysin, SNAP-25, and PSD-95 in brains from H67D/HFE expressing mice (Figure 4). Our findings are also consistent with studies demonstrating the need for cholesterol for synaptic vesicle recycling, regulation of synaptogenesis, and neurite outgrowth (Koudinov and Koudinova, 2005; Linetti et al., 2010).

Cholesterol is essential for proper myelination (Saher et al., 2005). Myelin production and maintenance play a strong role in cognitive function (Bartzokis et al., 2006; Bartzokis et al., 2007)). Cholesterol accounts for ~40% of the lipid content of myelin (O'Brien and Sampson, 1965). Therefore, a reduction in cholesterol in the brains of H67D-HFE mice would be expected to result in disruption of myelin maintenance. Consistent with that expectation, we observed a reduction in expression of proteins (Caspr and NogoA) involved in myelin compaction. Contactin-associated protein (Caspr) is an essential component of paranodal junctions (Einheber et al., 1997) and alterations in its expression and location have been proposed to affect myelin maintenance. Paranodal junctions are the sites where proteins interact to hold myelin fibers together and therefore, play an important role in myelin integrity (Coetzee et al., 1996). Another observation that supports impaired myelination in H67D-HFE mice is the elevation of iron in the brains of these animals (Figure 3E) as well as elevated oxidative stress and lipid peroxidation (Wint et a. unpublished data). Interestingly, increased lipid peroxidation has been associated with more myelin loss in older compared to younger individuals and in AD patients compared to normal older subjects (Bartzokis et al., 2003; Bartzokis et al., 2004; Chia et al., 1984; Peters and Sethares, 2002).

Cognitive function also requires appropriate maintenance of cholesterol homeostasis (Schreurs, 2010). There is a large body of data assessing the role of cholesterol in AD but

118 the relative contribution of brain and serum cholesterol to AD is controversial. Several studies show that high cholesterol in the plasma membrane promotes amyloid beta production (Refolo et al., 2001) and that inhibition of cholesterol synthesis improves memory in certain animal models (Kessler et al., 1986). On the other hand, some studies indicate that lowering brain cholesterol is detrimental to learning and memory (Endo et al., 1996; Xu et al., 1998) and that learning deficits could be reversed by feeding cholesterol (Dufour et al., 2006; Micale et al., 2008; Miller and Wehner, 1994; Xu et al., 1998). Genetic mutations that altered cholesterol levels in mice have also resulted in conflicting findings; some found that elevation of cholesterol resulted in better memory function while others reported that reducing cholesterol resulted in better learning and memory. A consistent observation in human studies is that higher levels of serum cholesterol are detrimental in middle age, but improve cognition in elderly individuals (Schreurs, 2010). In addition, lower total serum cholesterol (Kim et al., 2002) and lower HDL cholesterol were found to be associated with poor cognition (Atzmon et al., 2002). Consistent with the findings reported by Schreurs and Kim et al., we found higher amounts of serum cholesterol in 12 month old (middle age) and lower serum HDL cholesterol in 18 month old (older age) H67D-HFE mice (Fig. 11A & B). Our finding that the recognition and spatial learning and memory deficits seen in both 18 and 24 month old H67D-HFE mice correlated with their lower brain cholesterol content (Fig. 10) agrees with observations indicating that lower brain cholesterol levels impair learning and memory. Together, these observations underscore the importance of cholesterol for learning and memory and highlight the fact that excessive reduction in cholesterol levels within the brain can be accompanied by memory impairment.

Changes in regional and total brain volume have been described in a number of neurodevelopmental and neurodegenerative diseases in both humans and animal models of the human diseases. Recently, Nieman et al. (2007) reported that 13 transgenic mouse models with behavioral phenotypes had associated neuroanatomical abnormalities including volumetric changes as assessed by MRI that were consistent with the human phenotype. These observations suggest that there is a strong behavior-structure relationship (Nieman et al., 2007). In line with this we found that the volume of the mouse brains correlated significantly with measures of brain cholesterol content,

119 expression of synaptic proteins, and memory assessment. We found a 6.5% reduction in total brain volume of 24 month old H67D-HFE mice, as well as a significant enlargement of ventricles, and a reduction in volume of the corpus callosum. These observations are consistent with neuroanatomical abnormalities reported in MRI studies of brains of AD patients (Chan et al., 2003; Thompson et al., 2007). Specially, human AD subjects with the ApoE4 allele have significantly more loss of gray matter (Spampinato et al., 2011) and greater atrophy of the hippocampal formation (Pievani et al., 2011). Notably, older mice expressing human ApoE4 exhibit more brain atrophy and spatial memory deficits than age-matched WT-HFE mice (Yin et al., 2011). A significant volumetric change with a magnitude of 4-5% in brain has also been reported in mouse models of Huntington disease (Lerch et al., 2008) and Prader-Willi syndrome (Mercer et al., 2009) suggesting that this magnitude of volume change is associated with behavioral phenotype and could be clinically relevant.

Effect of iron on cholesterol metabolism

Our study suggested an inverse relationship between iron and cholesterol content in the HFE variant cells and in the brains of mice expressing WT- or H67D-HFE. Our findings are consistent with those of Brunet et al. (1997) who reported that treatment of rats with an iron-salicylate complex resulted in an elevation of iron, products of lipid peroxidation, and reduced total cholesterol in their serum. Treatment of rats with an iron dextran complex resulted in a decrease in total serum cholesterol regardless of whether the rats were fed a normal diet or one high in cholesterol (Turbino-Ribeiro et al., 2003). Studies in rabbits treated with iron dextran complex and fed a normal diet also showed similar effects (Dabbagh et al., 1997). This inverse relationship between iron and cholesterol content has also been observed in human studies. Individuals homozygous for C282Y- HFE, another variant of HFE that causes iron-overload, showed evidence of iron- overload and had a reduction in total and LDL cholesterol (Pankow et al., 2008). Recently, Du et al. (2012) showed an association between increased brain (nigrostriatal) iron content and low serum cholesterol in patients with Parkinson’s disease. One possible explanation for these effects is that they are mediated by iron-induced oxidative stress and lipid peroxidation. Excess iron has been shown to cause cellular injury by inducing

120 oxidative stress (Dabbagh et al., 1994) and elevation of lipid peroxidation (Britton et al., 1987). Depletion of energy (ATP and NADPH) due to oxidative stress and lipid- peroxidation mediated membrane damage has been shown to cause disruption of lipid synthesis and transport (Kehrer, 1993). In fact, diet-induced iron overload resulted in reduced expression of HMGCoAR and CYP7A1 (Brunet et al., 1999). Though most of these studies were focused on effects of excess iron on plasma cholesterol, it is possible that a similar mechanism exists in the brain. Collectively, these findings provide evidence for an inverse relationship between iron and cholesterol which merits further investigation.

In conclusion, we have shown that expression of H63D-HFE in cells and H67D-HFE in mice, was accompanied by a significant reduction in both cell and brain cholesterol. We also found that expression of H67D-HFE was associated with a reduction in expression of synaptic proteins and proteins needed for myelination, an increase in neurodegeneration, and impaired learning and memory. This is the first study to show that in addition to affecting iron homeostasis, a variant of HFE affects cholesterol metabolism and specifically links the H67D-HFE phenotype to decreases in learning and memory. Our observations also provide an insight into how H63D-HFE may contribute to earlier onset of AD in people expressing both ApoE4 and H63D-HFE.

Acknowledgements: We thank Dr. Erica Unger for using atomic absorption spectroscopy to measure iron, Dr. Kevin Boyd and Caesar Imperio for technical assistance and helpful feedback on behavior studies, Patti Miller and Zeinab Nasrullah for technical assistance with the MRI studies, Wint Nandar for providing brain lysates for 12 month old mice, Dr. Padmavathi Ponnuru for helpful discussion about iron treatment studies and Dr. Amanda Snyder for helpful feedback on fluorescent microscopy. This work was supported by the George M. Leader Laboratory for Alzheimer’s disease research and Harbolis AD research endowment.

121 Table 2.1: Expression of genes involved in cholesterol metabolism by H63D-HFE- expressing cells relative to WT-HFE-expressing cells. Pools of total RNA from WT- or H63D-HFE stably transfected cells were analyzed using the GEArray human target specific gene array. Changes in mRNA levels are expressed as  (fold increase),  (fold decrease). The average changes in expression (> 2 folds) are shown.

122 Table 2.2: Summary of cellular, biochemical, anatomical, and behavioral findings in H63D/ H67D-HFE relative to WT-HFE cells and mice

123 Figure 2.1: Effect of expression of H63D-HFE on cholesterol and cholesterol metabolizing enzymes. A) Cholesterol content of SH-SY5Y cells expressing WT- or H63D-HFE. Lipids were extracted from equal number of cells (106) of each cell type and cholesterol content measured. *** p<0.0005 (n=3). B) Filipin staining of WT- and H63D- HFE cells. C) Immunoblot analysis of cholesterol 24S-hydroxylase (CYP46A1) with actin as a loading control, and D) HMGCoA reductase (HMGCoAR) normalized to actin for WT- and H63D-HFE expressing cells. Cumulatively, these data support the hypothesis that expression of H63D-HFE is associated with decreased cholesterol synthesis and increased cholesterol catabolism, which could ultimately cause the reduction in total cholesterol shown in (A). E) Cholesterol content after treatment with 100uM FAC or 20uM DFO. Untreated controls served as controls. F) Immunoblot analysis of transferrin receptor (TfR) with actin as a loading control, showing iron status of cells treated with FAC or DFO.

124 Figure 2.2: Effect of cholesterol manipulation on survival of HFE-variant cells

A) Treatment of H63D-HFE expressing cells with 10 µM simvastatin resulted in a significant decrease in survival than seen for control WT-HFE cells treated analogously. ** P < 0.01, n=8). B) WT and H63D expressing cells were treated with 50µM of 24S- hydroxycholesterol (24S-HC; product of CYP46A1) for 28 hours (** p <0.01; n=8). Untreated cells served as control. Decreased cell survival was observed for H63D-HFE expressing cells relative to expressing WT-HFE after treatment with 24S-HC. These data show that reduction in cholesterol is also associated with decrease in survival of H63D cells and further decreases upon treatment with an inhibitor of cholesterol synthesis or a cholesterol metabolite.

125 Figure 2.3: Effect of expression of H67D-HFE on expression of cholesterol and cholesterol metabolizing enzymes in brains of mice. A) Total cholesterol content of WT- and H67D-HFE (homologous to human H63D) mouse brains at 12, 18, and 24 months (n=7-17 per group including both male and female mice). Lipids were extracted from 100µl of brain homogenate, cholesterol content was measured, and normalized to protein content. The observed, age-related decline in brain cholesterol was more significant in brains from H67D-HFE mice than those from WT- n=7-17). Immunoblot analyses of mouse brain B) cholesterol 24S-hydroxylase (CYP46A1), C) HMGCoA reductase (HMGCoAR), and D) low density lipoprotein receptor (LDLR), each normalized to actin (*p <0.05, **p<0.01, p>0.05, respectively; n=3-5). According to antibody manufacturer, two major bands indicate HMGCoAR. E) Total brain iron content of 18 month old WT- and H67D-HFE mice. Atomic absorption analysis of brains of 18 month old mice showed that they had higher levels of iron than those of WT-HFE mice. This is a paradigm shifting finding which along with recent human MRI studies [1, 2] indicate that, contrary to traditional paradigms, the brain is not protected from acquiring excess iron when the H67D-HFE mutation is present.

126

127 Figure 2.4: H67D-HFE alters expression of synaptic proteins. Immunoblot analyses of total expression of A) synaptophysin, B) SNAP-25, and C) PSD-95 in homogenates of brains from 18 month old mice (n=7-9 per group including both male and female mice; **p< 0.01, *** p<0.001). Actin was used as a loading control. Graphs show change in expression of synaptic proteins normalized to actin as a percentage of WT-HFE expression (control). According to antibody manufacturer, both bands represent synaptophysin. Densitometry analysis of both bands shows the same trend of change in expression between two genotypes. Decrease in the expression of proteins regulating the synaptic machinery suggests disruption of synaptic function in H67D-HFE mice.

128 Figure 2.5: Effect of H67D-HFE on expression of proteins involved in myelin maintenance. Immunoblot analysis of expression of A) NogoA, B) Caspr, and C) CNPase in homogenates of brains from 18 month old mice (n=7-9 per group including both male and female mice; *** p<0.001). Actin was used as a loading control. Graphs show change in expression of myelin proteins normalized to actin as a percentage of WT- HFE expression. The decrease in expression of proteins regulating myelin maintenance suggests disruption of myelination in H67D-HFE mice.

129 Figure 2.6: H67D-HFE induces neurodegeneration. Immunoblot analyses of total expression of neuronal marker A) NeuN, B) Total caspase-3, and C) cleaved caspase-3 in homogenates of brains from 18 month old mice (n=7-9 per group including both male and female mice; * p<0.05, *** p<0.001). Actin was used as a loading control. Graphs show change in expression of proteins normalized to actin as a percentage of WT-HFE expression (control). (D) Parietal cortex and (E) hippocampal CA3 regions of WT and H67D mouse brain sections (n = 6 mice/group) were coimmunostained with anti-NeuN (red) and anti-Cleaved Caspase-3 (green) to assess active caspase mediated apoptosis. Bars represent 20 µm. To confirm regional apoptotic differences, a terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) fluorometric assay was performed on parietal cortex (F) and hippocampal CA3 (G) regions of WT and H67D mouse brain sections (n = 6 mice/group). All cell nuclei were stained with propidium iodide (PI; red) and apoptotic cells indicated by TUNEL (green). Bars represent 50 µm (F) and 20 µm (G). (H) Apoptotic prevalence in parietal cortex (left panel) and pyramidal cells of hippocampus (right panel), as indicated by the yellow hue from TUNEL fluorescence and PI colocalization, was quantified with Image J cell counter and expressed as a percent of all cells positive for PI ± SEM.

130

131 Figure 2.7: Total brain volume (gray and white matter) of 22 month old WT- and H67D-HFE expressing mice. A) Representative image of brains of WT- and H67D-HFE mice showing enlarged lateral ventricles and thinner corpus callosum in the H67D brain, shown by arrow and arrow head, respectively. B) Quantitative analysis of brain regions. MRI was done on the brains of 22 month old mice, followed by manual segmentation and quantification of the entire brain, slice by slice, using MRIcro software. Results show a reduction in brain volume in H67D-HFE mice compared to brain volume in WT-HFE mice (n=10 per group; p-value=0.0165 for total brain volume using unpaired t-test). C) Total wet weight of the brains of mice when harvested after sacrifice. These data indicate that expression of H67D-HFE is associated with greater loss of brain volume than with expression of WT- HFE.

132 Figure 2.8: Effect of expression of H67D-HFE on learning and memory- Latency measures. A) Performance of H67D-HFE mice in the novel object recognition task showed that they had ~30% decrease in preference for the novel object than WT-HFE mice at both 14 and 24 months of age showing impairment of recognition memory (n=10-15 mice per group; ** p= <0.01, *** p= <0.001). B) Performance of H67D-HFE mice in the novel object location task, C) Spatial memory was assessed by the Barnes maze test. The time a mouse took to find the escape hole is defined as primary latency, while total latency is the time to enter the escape hole. Primary C) and total latency D) of 18-month old and 24 month old (F,G, respectively) H67D-HFE mice were significantly greater than those for WT-HFE mice during acquisition trials. Probe trials on days 5 and 12 which are tests of short and long term memory also show memory loss in E) 18 and H) 24 month old H67D-HFE mice (n=10-15 per group; ** p= <0.01, *** p= <0.001) . Significance indicated on all graphs shows an overall effect of genotype.

133 Figure 2.9: Effect of expression of H67D-HFE on learning and memory-Total Distance & errors. Total distance traveled on the Barnes maze for the first visit to escape hole was assessed using ANYmaze behavior tracking software. Both 18 and 24 month old H67D-HFE mice traveled longer distance to find the escape hole during acquisition trials (A,E) and probe trials (B, F). An error was defined as a nose poke in a hole other than the escape hole. Unlike 18 month old (C,D), 24 month old H67D mice made significantly more mistakes during acquisition trials (G) and probe trials (H).

134 Figure 2.10: Correlation analysis

135 Figure 2.11: Serum cholesterol content of WT- and H63D-HFE mice

A) Total serum cholesterol of 6, 12, and 18 month old mice

B) Comparison of total, HDL, and LDL cholesterol in serum of 18 month old mice, respectively

136 References

Adams, P.C., D.M. Reboussin, J.C. Barton, C.E. McLaren, J.H. Eckfeldt, G.D. McLaren, F.W. Dawkins, R.T. Acton, E.L. Harris, V.R. Gordeuk, C. Leiendecker-Foster, M. Speechley, B.M. Snively, J.L. Holup, E. Thomson, and P. Sholinsky. 2005. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med. 352:1769-78.

Ali-Rahmani, F., J.A. Hengst, J.R. Connor, and C.L. Schengrund. 2011. Effect of HFE Variants on Sphingolipid Expression by SH-SY5Y Human Neuroblastoma Cells. Neurochem Res.

Arosio, P., and S. Levi. 2002. Ferritin, iron homeostasis, and oxidative damage. Free Radical Biology and Medicine. 33:457-463.

Bartzokis, G., J.L. Cummings, D. Sultzer, V.W. Henderson, K.H. Nuechterlein, and J. Mintz. 2003. White matter structural integrity in healthy aging adults and patients with Alzheimer disease - A magnetic resonance imaging study. Archives of Neurology. 60:393-398.

Bartzokis, G., P.H. Lu, D.H. Geschwind, N. Edwards, J. Mintz, and J.L. Cummings. 2006. Apolipoprotein E genotype and age-related myelin breakdown in healthy individuals: implications for cognitive decline and dementia. Arch Gen Psychiatry. 63:63-72.

Bartzokis, G., P.H. Lu, D.H. Geschwind, K. Tingus, D. Huang, M.F. Mendez, N. Edwards, and J. Mintz. 2007. Apolipoprotein E affects both myelin breakdown and cognition: implications for age-related trajectories of decline into dementia. Biol Psychiatry. 62:1380-7.

137 Bartzokis, G., P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P. Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely, and J.R. Connor. Prevalent iron metabolism gene variants associated with increased brain ferritin iron in healthy older men. J Alzheimers Dis. 20:333-41.

Bartzokis, G., D. Sultzer, P.H. Lu, K.H. Nuechterlein, J. Mintz, and J.L. Cummings. 2004. Heterogeneous age-related breakdown of white matter structural integrity: implications for cortical "disconnection" in aging and Alzheimer's disease. Neurobiol Aging. 25:843-51.

Berg, D., U. Hoggenmuller, E. Hofmann, R. Fischer, M. Kraus, M. Scheurlen, and G. Becker. 2000. The basal ganglia in haemochromatosis. Neuroradiology. 42:9-13.

Britton, R.S., B.R. Bacon, and R.O. Recknagel. 1987. Lipid-Peroxidation and Associated Hepatic Organelle Dysfunction in Iron Overload. Chemistry and Physics of Lipids. 45:207-239.

Brunet, S., F. Guertin, L. Thibault, V. Gavino, E. Delvin, and E. Levy. 1997. Iron- salicylate complex induces peroxidation, alters hepatic lipid profile and affects plasma lipoprotein composition. Atherosclerosis. 129:159-168.

Brunet, S., L. Thibault, E. Delvin, W. Yotov, M. Bendayan, and E. Levy. 1999. Dietary iron overload and induced lipid peroxidation are associated with impaired plasma lipid transport and hepatic sterol metabolism in rats. Hepatology. 29:1809-1817.

Bu, G. 2009. Apolipoprotein E and its receptors in Alzheimer's disease: pathways, pathogenesis and therapy. Nat Rev Neurosci. 10:333-44.

Burt, M.J., P.M. George, J.D. Upton, J.A. Collett, C.M. Frampton, T.M. Chapman, T.A. Walmsley, and B.A. Chapman. 1998. The significance of haemochromatosis gene mutations in the general population: implications for screening. Gut. 43:830-6.

138 Bush, A.I. 2003. The metallobiology of Alzheimer's disease. Trends Neurosci. 26:207-14.

Butterfield, D.A., E. Barone, F. Di Domenico, G. Cenini, R. Sultana, M.P. Murphy, C. Mancuso, and E. Head. Atorvastatin treatment in a dog preclinical model of Alzheimer's disease leads to up-regulation of haem oxygenase-1 and is associated with reduced oxidative stress in brain. Int J Neuropsychopharmacol:1-7.

Cartagena, C.M., M.P. Burns, and G.W. Rebeck. 2010. 24S-hydroxycholesterol effects on lipid metabolism genes are modeled in traumatic brain injury. Brain Res. 1319:1-12.

Chan, D., J.C. Janssen, J.L. Whitwell, H.C. Watt, R. Jenkins, C. Frost, M.N. Rossor, and N.C. Fox. 2003. Change in rates of cerebral atrophy over time in early-onset Alzheimer's disease: longitudinal MRI study. Lancet. 362:1121-1122.

Chia, L.S., J.E. Thompson, and M.A. Moscarello. 1984. X-Ray-Diffraction Evidence for Myelin Disorder in Brain from Humans with Alzheimers-Disease. Biochimica Et Biophysica Acta. 775:308-312.

Coetzee, T., N. Fujita, J. Dupree, R. Shi, A. Blight, K. Suzuki, and B. Popko. 1996. Myelination in the absence of galactocerebroside and sulfatide: normal structure with abnormal function and regional instability. Cell. 86:209-19.

Connor, J.R., and S.Y. Lee. 2006. HFE mutations and Alzheimer's disease. J Alzheimers Dis. 10:267-76.

Connor, J.R., S.L. Menzies, S.M. St Martin, and E.J. Mufson. 1992. A histochemical study of iron, transferrin, and ferritin in Alzheimer's diseased brains. J Neurosci Res. 31:75-83.

139 Dabbagh, A.J., T. Mannion, S.M. Lynch, and B. Frei. 1994. The Effect of Iron Overload on Rat Plasma and Liver Oxidant Status in-Vivo. Biochemical Journal. 300:799- 803.

Dabbagh, A.J., G.T. Shwaery, J.F. Keaney, and B. Frei. 1997. Effect of iron overload and iron deficiency on atherosclerosis in the hypercholesterolemic rabbit. Arteriosclerosis Thrombosis and Vascular Biology. 17:2638-2645.

Dietschy, J.M., and S.D. Turley. 2001. Cholesterol metabolism in the brain. Curr Opin Lipidol. 12:105-12.

Du, G.W., M.M. Lewis, M.L. Shaffer, H.L. Chen, Q.X. Yang, R.B. Mailman, and X.M. Huang. 2012. Serum Cholesterol and Nigrostriatal R2*Values in Parkinson's Disease. PLoS One. 7.

Dufour, F., Q.Y. Liu, P. Gusev, D. Alkon, and M. Atzori. 2006. Cholesterol-enriched diet affects spatial learning and synaptic function in hippocampal synapses. Brain Res. 1103:88-98.

Einheber, S., G. Zanazzi, W. Ching, S. Scherer, T.A. Milner, E. Peles, and J.L. Salzer. 1997. The axonal membrane protein Caspr, a homologue of neurexin IV, is a component of the septate-like paranodal junctions that assemble during myelination. J Cell Biol. 139:1495-506.

Endo, Y., J.I. Nishimura, and F. Kimura. 1996. Impairment of maze learning in rats following long-term glucocorticoid treatments. Neurosci Lett. 203:199-202.

Fang, B.Y., and J.P. Jia. 2007. [The effect of two newly Chinese presenilin-1 mutations on the sensitivity to trophic factor withdrawal in human neuroblastoma cells]. Zhonghua Yi Xue Za Zhi. 87:336-40.

Fuhrman, B., J. Oiknine, and M. Aviram. 1994. Iron induces lipid peroxidation in cultured macrophages, increases their ability to oxidatively modify LDL, and affects their secretory properties. Atherosclerosis. 111:65-78.

140 Gellermann, G.P., T.R. Appel, A. Tannert, A. Radestock, P. Hortschansky, V. Schroeckh, C. Leisner, T. Lutkepohl, S. Shtrasburg, C. Rocken, M. Pras, R.P. Linke, S. Diekmann, and M. Fandrich. 2005. Raft lipids as common components of human extracellular amyloid fibrils. Proc Natl Acad Sci U S A. 102:6297-302.

Ginsberg, L., J.R. Atack, S.I. Rapoport, and N.L. Gershfeld. 1993a. Evidence for a membrane lipid defect in Alzheimer disease. Mol Chem Neuropathol. 19:37-46.

Ginsberg, L., J.R. Atack, S.I. Rapoport, and N.L. Gershfeld. 1993b. Regional specificity of membrane instability in Alzheimer's disease brain. Brain Res. 615:355-7.

Gong, J.-S., M. Kobayashi, H. Hayashi, K. Zou, N. Sawamura, S.C. Fujita, K. Yanagisawa, and M. Michikawa. 2002. Apolipoprotein E (ApoE) Isoform- dependent Lipid Release from Astrocytes Prepared from Human ApoE3 and ApoE4 Knock-in Mice. J. Biol. Chem. 277:29919-29926.

Hardie, D.G. 2003. Minireview: the AMP-activated protein kinase cascade: the key sensor of cellular energy status. Endocrinology. 144:5179-83.

Kanayama, T., M. Arito, K. So, S. Hachimura, J. Inoue, and R. Sato. 2007. Interaction between sterol regulatory element-binding proteins and liver receptor homolog-1 reciprocally suppresses their transcriptional activities. Journal of Biological Chemistry. 282:10290-10298.

Kawano, M., M. Kawakami, M. Otsuka, H. Yashima, T. Yaginuma, and A. Ueki. 1995. Marked decrease of plasma apolipoprotein AI and AII in Japanese patients with late-onset non-familial Alzheimer's disease. Clin Chim Acta. 239:209-11.

Kehrer, J.P. 1993. Free-Radicals as Mediators of Tissue-Injury and Disease. Critical Reviews in Toxicology. 23:21-48.

141 Kenney, J.W., M.D. Adoff, D.S. Wilkinson, and T.J. Gould. 2011. The effects of acute, chronic, and withdrawal from chronic nicotine on novel and spatial object recognition in male C57BL/6J mice. Psychopharmacology (Berl). 217:353-65.

Kessler, A.R., B. Kessler, and S. Yehuda. 1986. In vivo modulation of brain cholesterol level and learning performance by a novel plant lipid: indications for interactions between hippocampal-cortical cholesterol and learning. Life Sci. 38:1185-92.

Kim, J.M., R. Stewart, I.S. Shin, and J.S. Yoon. 2002. Low cholesterol, cognitive function and Alzheimer s disease in a community population with cognitive impairment. J Nutr Health Aging. 6:320-3.

Kirsch, C., G.P. Eckert, and W.E. Mueller. 2003. Statin effects on cholesterol microdomains in brain plasma membranes. Biochem Pharmacol. 65:843-56.

Koudinov, A.R., and N.V. Koudinova. 2005. Cholesterol homeostasis failure as a unifying cause of synaptic degeneration. J Neurol Sci. 229-230:233-40.

Lambert, J.C., D. Mann, F. Richard, J. Tian, J. Shi, U. Thaker, S. Merrot, J. Harris, B. Frigard, T. Iwatsubo, C. Lendon, and P. Amouyel. 2005. Is there a relation between APOE expression and brain amyloid load in Alzheimer's disease? Journal of Neurology Neurosurgery and Psychiatry. 76:928-933.

Ledesma, M.D. 2003. Raft disorganization leads to reduced plasmin activity in Alzheimer's disease brains. EMBO Rep. 4:1190-1196.

Ledesma, M.D., and C.G. Dotti. 2006. Amyloid excess in Alzheimer's disease: what is cholesterol to be blamed for? FEBS Lett. 580:5525-5532.

Lee, S.Y., S.M. Patton, R.J. Henderson, and J.R. Connor. 2007. Consequences of expressing mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking endogenous HFE. FASEB J. 21:564-76.

Leoni, V. 2009. Oxysterols as markers of neurological disease--a review. Scand J Clin Lab Invest. 69:22-5.

142 Lerch, J.P., J.B. Carroll, S. Spring, L.N. Bertram, C. Schwab, M.R. Hayden, and R.M. Henkelman. 2008. Automated deformation analysis in the YAC128 Huntington disease mouse model. Neuroimage. 39:32-39.

Liao, W.L., G.Y. Heo, N.G. Dodder, R.E. Reem, N. Mast, S. Huang, P.L. Dipatre, I.V. Turko, and I.A. Pikuleva. 2010. Quantification of cholesterol-metabolizing P450s CYP27A1 and CYP46A1 in neural tissues reveals a lack of enzyme-product correlations in human retina but not human brain. J Proteome Res. 10:241-8.

Linetti, A., A. Fratangeli, E. Taverna, P. Valnegri, M. Francolini, V. Cappello, M. Matteoli, M. Passafaro, and P. Rosa. 2010. Cholesterol reduction impairs exocytosis of synaptic vesicles. J Cell Sci. 123:595-605.

Liu, G., M.R. Garrett, P. Men, X. Zhu, G. Perry, and M.A. Smith. 2005. Nanoparticle and other metal chelation therapeutics in Alzheimer disease. Biochim Biophys Acta. 1741:246-52.

Liu, G., W. Huang, R.D. Moir, C.R. Vanderburg, B. Lai, Z. Peng, R.E. Tanzi, J.T. Rogers, and X. Huang. 2006. Metal exposure and Alzheimer's pathogenesis. J Struct Biol. 155:45-51.

Lopez-Lopez, C., M.O. Dietrich, F. Metzger, H. Loetscher, and I. Torres-Aleman. 2007. Disturbed cross talk between insulin-like growth factor I and AMP-activated protein kinase as a possible cause of vascular dysfunction in the amyloid precursor protein/presenilin 2 mouse model of Alzheimer's disease. J Neurosci. 27:824-31.

143 Lukiw, W.J. 2006. Cholesterol and 24S-hydroxycholesterol trafficking in Alzheimer's disease. Expert Rev Neurother. 6:683-93.

Lutjohann, D. 2006. Cholesterol metabolism in the brain: importance of 24S- hydroxylation. Acta Neurol Scand Suppl. 185:33-42.

Lutjohann, D., M. Stroick, T. Bertsch, S. Kuhl, B. Lindenthal, K. Thelen, U. Andersson, I. Bjorkhem, K. Bergmann Kv, and K. Fassbender. 2004. High doses of simvastatin, pravastatin, and cholesterol reduce brain cholesterol synthesis in guinea pigs. Steroids. 69:431-8.

Markesbery, W.R. 1997. Oxidative stress hypothesis in Alzheimer's disease. Free Radic Biol Med. 23:134-47.

Maynard, C.J., A.I. Bush, C.L. Masters, R. Cappai, and Q.X. Li. 2005. Metals and amyloid-beta in Alzheimer's disease. Int J Exp Pathol. 86:147-59.

Mercer, R.E., E.M. Kwolek, J.M. Bischof, M. van Eede, R.M. Henkelman, and R. Wevrick. 2009. Regionally Reduced Brain Volume, Altered Serotonin Neurochemistry, and Abnormal Behavior in Mice Null for the Circadian Rhythm Output Gene Magel2. American Journal of Medical Genetics Part B- Neuropsychiatric Genetics. 150B:1085-1099.

Micale, V., G. Scapagnini, C. Colombrita, C. Mazzola, D.L. Alkon, and F. Drago. 2008. Behavioral effects of dietary cholesterol in rats tested in experimental models of mild stress and cognition tasks. Eur Neuropsychopharmacol. 18:462-71.

Millan, J., X. Pinto, A. Munoz, M. Zuniga, J. Rubies-Prat, L.F. Pallardo, L. Masana, A. Mangas, A. Hernandez-Mijares, P. Gonzalez-Santos, J.F. Ascaso, and J. Pedro-

144 Botet. 2009. Lipoprotein ratios: Physiological significance and clinical usefulness in cardiovascular prevention. Vasc Health Risk Manag. 5:757-65.

Miller, R.G. 1984. The Use and Abuse of Filipin to Localize Cholesterol in Membranes. Cell Biology International Reports. 8:519-535.

Miller, S., and J.M. Wehner. 1994. Cholesterol Treatment Facilitates Spatial-Learning Performance in Dba/2ibg Mice. Pharmacology Biochemistry and Behavior. 49:257-261.

Miron, V.E., S.P. Zehntner, T. Kuhlmann, S.K. Ludwin, T. Owens, T.E. Kennedy, B.J. Bedell, and J.P. Antel. 2009. Statin therapy inhibits remyelination in the central nervous system. Am J Pathol. 174:1880-90.

Mitchell, R.M., S.Y. Lee, Z. Simmons, and J.R. Connor. 2010. HFE polymorphisms affect cellular glutamate regulation. Neurobiology of Aging. In Press, Corrected Proof.

Nandar, W., and J.R. Connor. HFE gene variants affect iron in the brain. J Nutr. 141:729S-739S.

Nieman, B.J., J.P. Lerch, N.A. Bock, X.J. Chen, J.G. Sled, and R.M. Henkelman. 2007. Mouse behavioral mutants have neuroimaging abnormalities. Human Brain Mapping. 28:567-575.

145 O'Brien, J.S., and E.L. Sampson. 1965. Lipid composition of the normal human brain: gray matter, white matter, and myelin. J Lipid Res. 6:537-44.

Okada, S. 1996. Iron-induced tissue damage and cancer: the role of reactive oxygen species-free radicals. Pathol Int. 46:311-32.

Pankow, J.S., E. Boerwinkle, P.C. Adams, E. Guallar, C. Leiendecker-Foster, J. Rogowski, and J.H. Eckfeldt. 2008. HFE C282Y homozygotes have reduced low- density lipoprotein cholesterol: the Atherosclerosis Risk in Communities (ARIC) Study. Translational Research. 152:3-10.

Papassotiropoulos, A., D. Lutjohann, M. Bagli, S. Locatelli, F. Jessen, R. Buschfort, U. Ptok, I. Bjorkhem, K. von Bergmann, and R. Heun. 2002. 24S-hydroxycholesterol in cerebrospinal fluid is elevated in early stages of dementia. J Psychiatr Res. 36:27-32.

Patil, S.S., B. Sunyer, H. Hoger, and G. Lubec. 2009. Evaluation of spatial memory of C57BL/6J and CD1 mice in the Barnes maze, the Multiple T-maze and in the Morris water maze. Behav Brain Res. 198:58-68.

Percy, M., S. Moalem, A. Garcia, M.J. Somerville, M. Hicks, D. Andrews, A. Azad, P. Schwarz, R.B. Zavareh, R. Birkan, C. Choo, V. Chow, S. Dhaliwal, V. Duda, A.L. Kupferschmidt, K. Lam, D. Lightman, K. Machalek, W. Mar, F. Nguyen, P.J. Rytwinski, E. Svara, M. Tran, K. Wheeler, L. Yeung, K. Zanibbi, R. Zener, M. Ziraldo, and M. Freedman. 2008. Involvement of ApoE E4 and H63D in sporadic Alzheimer's disease in a folate-supplemented Ontario population. J Alzheimers Dis. 14:69-84.

Peters, A., and C. Sethares. 2002. Aging and the myelinated fibers in prefrontal cortex and corpus callosum of the monkey. Journal of Comparative Neurology. 442:277- 291.

Pievani, M., S. Galluzzi, P. Thompson, P.E. Rasser, M. Bonetti, and G. Frisoni. 2011. APOE4 Genotype Is Associated with Greater Atrophy of the Hippocampal Formation in Alzheimer's Disease. Neurology. 76:A300-A300.

146 Pulliam, J.F., C.D. Jennings, R.J. Kryscio, D.G. Davis, D. Wilson, T.J. Montine, F.A. Schmitt, and W.R. Markesbery. 2003. Association of HFE mutations with neurodegeneration and oxidative stress in Alzheimer's disease and correlation with APOE. Am J Med Genet B Neuropsychiatr Genet. 119B:48-53.

Refolo, L.M., M.A. Pappolla, J. LaFrancois, B. Malester, S.D. Schmidt, T. Thomas- Bryant, G.S. Tint, R. Wang, M. Mercken, S.S. Petanceska, and K.E. Duff. 2001. A cholesterol-lowering drug reduces beta-amyloid pathology in a transgenic mouse model of Alzheimer's disease. Neurobiology of Disease. 8:890-899.

Roher, A.E., Y.M. Kuo, K.M. Kokjohn, M.R. Emmerling, and S. Gracon. 1999. Amyloid and lipids in the pathology of Alzheimer disease. Amyloid. 6:136-45.

Sampietro, M., L. Caputo, A. Casatta, M. Meregalli, A. Pellagatti, J. Tagliabue, G. Annoni, and C. Vergani. 2001. The hemochromatosis gene affects the age of onset of sporadic Alzheimer's disease. Neurobiol Aging. 22:563-8.

Schreurs, B.G. 2010. The effects of cholesterol on learning and memory. Neurosci Biobehav Rev. 34:1366-79.

Shobab, L.A., G.Y.R. Hsiung, and H.H. Feldman. 2005. Cholesterol in Alzheimer's disease. Lancet Neurol. 4:841-852.

Sjogren, M., M. Mielke, D. Gustafson, P. Zandi, and I. Skoog. 2006. Cholesterol and Alzheimer's disease--is there a relation? Mech Ageing Dev. 127:138-47.

Song, F., A. Poljak, J. Crawford, N.A. Kochan, W. Wen, B. Cameron, O. Lux, H. Brodaty, K. Mather, G.A. Smythe, and P.S. Sachdev. 2012. Plasma apolipoprotein levels are associated with cognitive status and decline in a community cohort of older individuals. PLoS One. 7:e34078.

147 Sottocornola, E., R. Misasi, V. Mattei, L. Ciarlo, R. Gradini, T. Garofalo, B. Berra, I. Colombo, and M. Sorice. 2006. Role of gangliosides in the association of ErbB2 with lipid rafts in mammary epithelial HC11 cells. FEBS J. 273:1821-30.

Spampinato, M.V., Z. Rumboldt, R.J. Hosker, J.E. Mintzer, and A.s.D. Neuroimaging. 2011. Apolipoprotein E and Gray Matter Volume Loss in Patients with Mild Cognitive Impairment and Alzheimer Disease. Radiology. 258:843-852.

Tan, T.C., D.H. Crawford, L.A. Jaskowski, T.M. Murphy, M.L. Heritage, V.N. Subramaniam, A.D. Clouston, G.J. Anderson, and L.M. Fletcher. 2011. Altered lipid metabolism in Hfe-knockout mice promotes severe NAFLD and early fibrosis. Am J Physiol Gastrointest Liver Physiol. 301:G865-76.

Thelen, K.M., P. Falkai, T.A. Bayer, and D. Lutjohann. 2006a. Cholesterol synthesis rate in human hippocampus declines with aging. Neurosci Lett. 403:15-9.

Thelen, K.M., K.M. Rentsch, U. Gutteck, M. Heverin, M. Olin, U. Andersson, A. von Eckardstein, I. Bjorkhem, and D. Lutjohann. 2006b. Brain cholesterol synthesis in mice is affected by high dose of simvastatin but not of pravastatin. J Pharmacol Exp Ther. 316:1146-52.

Thiele, C., M.J. Hannah, F. Fahrenholz, and W.B. Huttner. 2000. Cholesterol binds to synaptophysin and is required for biogenesis of synaptic vesicles. Nat Cell Biol. 2:42-9.

Thompson, P.M., K.M. Hayashi, R.A. Dutton, M.C. Chiang, A.D. Leow, E.R. Sowell, G. De Zubicaray, J.T. Becker, O.L. Lopez, H.J. Aizenstein, and A.W. Toga. 2007. Tracking Alzheimer's disease. Imaging and the Aging Brain. 1097:183-214.

Tian, G.L., Q.M. Kong, L.C. Lai, A. Ray-Chaudhury, and C.L.G. Lin. 2010. Increased expression of cholesterol 24S-hydroxylase results in disruption of glial glutamate transporter EAAT2 association with lipid rafts: a potential role in Alzheimer's disease. Journal of Neurochemistry. 113:978-989.

148 Tomatsu, S., K.O. Orii, R.E. Fleming, C.C. Holden, A. Waheed, R.S. Britton, M.A. Gutierrez, S. Velez-Castrillon, B.R. Bacon, and W.S. Sly. 2003. Contribution of the H63D mutation in HFE to murine hereditary hemochromatosis. Proc Natl Acad Sci U S A. 100:15788-93.

Turbino-Ribeiro, S.M.L., M.E. Silva, D.A. Chianca, H. de Paula, L.M. Cardoso, E. Colombari, and M.L. Pedrosa. 2003. Iron overload in hypercholesterolemic rats affects iron homeostasis and serum lipids but not blood pressure. Journal of Nutrition. 133:15-20.

Van Der Putten, C., H.F. Kuipers, E.A. Zuiderwijk-Sick, L. Van Straalen, I. Kondova, P.J. Van Den Elsen, and J.J. Bajramovic. 2011. Statins amplify TLR-induced responses in microglia via inhibition of cholesterol biosynthesis. Glia. 60:43-52.

Wu, S.B., and Y.H. Wei. 2012. AMPK-mediated increase of glycolysis as an adaptive response to oxidative stress in human cells: Implication of the cell survival in mitochondrial diseases. Biochimica Et Biophysica Acta-Molecular Basis of Disease. 1822:233-247.

Xu, G.R., R.J. Servatius, S. Shefer, G.S. Tint, W.T. O'Brien, A.K. Batta, and G. Salen. 1998. Relationship between abnormal cholesterol synthesis and retarded learning in rats. Metabolism-Clinical and Experimental. 47:878-882.

Yamanaka, K., Y. Saito, T. Yamamori, Y. Urano, and N. Noguchi. 2011. 24(S)- hydroxycholesterol induces neuronal cell death through necroptosis, a form of programmed necrosis. J Biol Chem. 286:24666-73.

Yates, M.A., and J.M. Juraska. 2007. Increases in size and myelination of the rat corpus callosum during adulthood are maintained into old age. Brain Research. 1142:13- 18.

149 Yin, J.X., G.H. Turner, H.J. Lin, S.W. Coons, and J. Shi. 2011. Deficits in Spatial Learning and Memory is Associated with Hippocampal Volume Loss in Aged Apolipoprotein E4 Mice. Journal of Alzheimers Disease. 27:89-98.

150 Chapter 3

Effect of HFE variants on sphingolipid expression by SH-SY5Y human neuroblastoma cells

Abstract

C282Y and H63D are two common variants of the hemochromatosis protein HFE. SH- SY5Y human neuroblastoma cells stably transfected to express either wild type HFE (WT-HFE), C282Y or H63D allele were analyzed for effect of expression of the mutant proteins on transcription of 14 enzymes involved in sphingolipid metabolism. Cells expressing the C282Y variant showed significant increases (>2-fold) in transcription of five genes and decreases in two compared to that seen for cells expressing WT-HFE, while cells expressing the H63D variant showed an elevation in transcription of one gene and a decrease in two. These changes were seen as alterations in ganglioside composition, cell surface binding by the binding subunit of cholera toxin, expression of sphingosine-kinase-1 and synthesis of sphingosine-1-phosphate. These changes may explain why C282Y-HFE is a risk factor for colon and breast cancer and possibly protective against Alzheimer’s disease while H63D-HFE is a risk factor for neurodegenerative diseases.

Introduction

The hemochromatosis gene product, HFE, helps regulate iron uptake into cells by interacting with transferrin receptors (TfR)1 and 2. TfR1 functions in cellular iron uptake while TfR2 helps regulate whole-body iron homeostasis (1). Mutations in HFE (H63D) have been independently implicated as risk factors for neurodegenerative diseases such as Alzheimer’s (AD), Parkinson’s, and motor neuron disease (2,3) and for breast and

151 colorectal cancer [(C282Y) 4]. In contrast to the H63D mutation being a risk factor for neurodegenerative diseases, the C282Y mutation was recently reported to have a negative association with AD, supporting a possible protective role for it in neurodegeneration (5).

Retention of the C282Y variant of HFE in the ER (6), results in induction of specific ER stress responses that can lead to apoptosis (7). In contrast the H63D variant is found on the cell surface where it can form stable complexes with the transferrin receptor (TfR) but fails to limit transferrin binding to the receptor (8). We found that cell surface HFE is associated with lipid rafts (see below), as is TfR2 (1,9) with which it interacts (10). Lipid rafts are discrete areas in the membrane enriched in sphingolipids and cholesterol relative to phospholipids as well as proteins involved in signal transduction (11). Specific mutations in a raft-associated protein such as HFE may alter its interaction with other raft components thereby disrupting signal transduction.

Historically it has been found that when cells undergo transformation (spontaneous or induced) they tend to have alterations in their glycosphingolipid (GSL) composition (eg. 12-14). More recently, we observed (15) that disruption of lipid rafts can affect subcellular distribution of the ganglioside GM1, a lipid raft marker. Interestingly, gene expression of ceramide kinase, galactosyl ceramide synthase, and ganglioside GD3 synthase were reported to be significantly altered in carriers of C282Y-HFE (16). In addition to accompanying the changes in regulation of cell growth seen upon transformation, glycosphingolipids have been shown to affect neurotransmitter receptor conformation and function (17), signal transduction (eg. 18,19), and accumulation of amyloid-ß-protein in Alzheimer’s disease (20-22). While gangliosides have been shown to induce assembly of amyloid ß-proteins (20,21), inhibition of glycosphingolipid synthesis was found to reduce secretion of both the ß-amyloid precursor protein and amyloid ß-peptide (23). Studies with SH-SY5Y human neuroblastoma cells indicated that enrichment of membrane cholesterol prevented association of Aß1-42 oligomers with GM1 (24). These observations coupled with our observation (unpublished) that SH-SY5Y

152 human neuroblastoma cells expressing the H63D mutant of HFE had reduced levels of cholesterol led us to investigate the sphingolipid composition of cells stably transfected to express either WT-, H63D- or C282Y-HFE.

After establishing that WT-HFE was associated with lipid rafts, we determined whether the H63D or C282Y variants of HFE induced changes in transcription of 14 genes involved in the metabolism of sphingolipids since they are enriched in lipid rafts. We found that cells expressing either the H63D or C282Y mutation transcribed specific genes differently than cells expressing WT-HFE and also had altered expression of specific sphingolipids.

Experimental procedures

Materials: SH-SY5Y human neuroblastoma cells, stably transfected (25,26) to express either FLAG-tagged wild-type-, H63D-, or C282Y-HFE alleles, were grown in DMEM/F12 containing 10% fetal bovine serum, 1% antibiotics (Penn-Strep, glutamine), non-essential amino acids, and 1.8g/L NaHCO3 at 37˚C in an atmosphere of 5% CO2. Alexa-Fluor® 594-conjugated cholera toxin binding subunit (CTxB, C-34777) was purchased from Molecular Probes (Eugene, OR), anti-sphingosine kinase 1 monoclonal antibody from Abnova [(MO1), Clone 1D6, Cat # H00008877-M01)], flotillin-1 mouse monoclonal antibody (610820) from BD Transduction Laboratories (San Jose, CA), and anti-FLAG monoclonal antibody (F3165), anti-actin monoclonal antibody (A5441), HRP-conjugated goat anti-mouse IgG (A4416), and 4’,6-diamidino-2-phenylindole (DAPI, D9542) were from Sigma (St. Louis, MO). Silica gel 60 high performance thin layer chromatography (HPTLC) plates were from VWR (Bridgeport, NJ), D-erythro- sphingosine from Avanti Polar Lipids (Alabaster, AL), ganglioside standards from Matreya (Pleasant Gap, PA), collagen coated chamber slides from BD Biosciences (San Jose, CA), and Halt Protease Inhibitors™ from PPierce (Rockford, IL). The RNeasy Mini, RT2 First Strand Kit, RT2 SYBR® Green/ROX™ qPCR Master Mix Kit, and the custom PCR array plate (CAPA1624-6) were obtained from SABiosciences, now part of Qiagen (Valencia, CA). 10% pre-cast gels were purchased from Bio-Rad (Hercules, CA)

153 - 32P ATP was obtained from New England Nuclear Cororation (part of DuPont).

Growth curves: Cells were seeded at 6X105 cells per well in six well plates and grown as described above. Starting at 72hr after seeding cells were harvested every twelve hrs for the time shown (Figure 1) and trypan blue negative cells counted. Phase contrast photomicrographs were taken 132 hr after seeding (Figure 1).

Isolation of lipid rafts: Because HFE has been found in association with TfR2, a raft associated protein, we determined whether WT-HFE was also associated with them. Rafts were isolated from cells transfected to express WT-HFE essentially as described by Petro et al (27). In brief, harvested cells were suspended in 25mM Tris-HCl, pH7.5, containing 150mM NaCl, 5mM EDTA, 1% Triton-X-100 and a 1:100 dilution of Halt Protease Inhibitors™. After incubation at 4˚C for 30min, the solution was added to an equal volume of 80% sucrose in 25mM Tris-HCl, pH 7.5 containing 150mM NaCl and 5mM EDTA (TNE) and overlaid first with 30% sucrose and then with 5% sucrose, both in TNE. Gradients were centrifuged at 4˚C in a Beckman SW41Ti for 19hr at 38,000RPM. One ml fractions were collected from the top of the gradient and proteins present in an equal volume of each fraction separated by SDS-PAGE, transferred to a PVDF membrane, and probed with either an anti-flotillin mouse monoclonal or anti- FLAG mouse monoclonal antibody, followed by exposure to a HRP-conjugated goat anti-mouse secondary antibody. Bands were visualized using SuperSignal West Femto Maximum Sensitivity Substrate™ and exposure in a SynGene Gnome (Frederick, MD).

Immunohistochemistry: Cells were seeded on collagen coated 4 well chamber slides , maintained as described previously and taken for analysis when WT cells were just confluent. Analyses were done as previously described (15). In brief, cells were washed with phosphate buffered saline (PBS) prior to fixation in 4% paraformaldehyde in PBS

154 (20 min, 4˚C). After blocking nonspecific binding sites [PBS-normal goat serum (9:1, v:v), 1 hr, room temperature] cells were washed with PBS, and exposed to CTxB-594 (5µg/ml) in PBS-NGS (200:3, v:v) overnight at 4°C. After washing to remove unbound CTxB-594, cells were visualized using a Nikon Eclipse 80i microscope connected to a Nikon Digital Sight camera and images captured using Nikon’s NIS-Elements Br Software,version 3.00.

Ganglioside analyses: Cells were harvested when just confluent and an aliquot taken for protein analysis using the Bio-Rad DC assay with BSA as the standard. Lipids were extracted from the remainder of the cells using the procedure described by Kasperzyk et al (28). In brief, cells were extracted with a 10-fold volume of chloroform-methanol (1:1, v:v), acidic lipids separated from neutral ones by chromatography on DEAE-Sephadex, (A-25), the acidic fraction dried, treated with mild base, and salts removed by chromatography on a C18 Bond Elut® column (Varian Inc., USA). Gangliosides were identified by HPTLC. Plates were developed using chloroform:2-propanol:50mM KCl

(2:13.4:4.6, v/v/v) or chloroform:methanol:0.3% CaCl2•2H2O (60:35:8, v/v/v) and sialic acid containing bands visualized using resorcinol spray (29). Resorcinol positive bands were identified by comparison of their migration with that of known ganglioside standards.

Gene expression: Total RNA was isolated from SH-SY5Y cells expressing either WT-, H63D-, or C282Y-HFE using the RNeasy Mini Kit according to the manufacturer’s instructions. Recovered RNA was quantified by its absorbance at 260nm. cDNA was prepared by reverse transcription of isolated RNA using the RT2 First Strand Kit. For qRT-PCR reactions, ~4ng cDNA samples were added to a commercially provided 384 well custom array containing primer pairs specific for the 14 genes selected (SABiosciences, Frederick, MD) and reactions were carried out according to directions provided with the RT2 SYBR® Green/ROX™ qPCR Master Mix Kit. Gene expression of

155 each of the 14 genes was monitored using an ABI 7800 (Applied Biosystems, USA). Data was normalized to expression of mRNA for both actin and ribosomal protein L13A.

Expression of sphingosine kinase 1 was monitored using Western blot analysis. Proteins present in cell samples containing 50µg of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on Bio-Rad 10% pre-cast gels, and then blotted to a PVDF membrane. Following transfer, non-specific binding blocked by incubating blots in Tris-buffered saline with 0.1% Tween, pH7.4 (TBST), containing 5% nonfat dried milk for an hour at room temperature. The blot was then probed using the anti-sphingosine kinase 1 antibody followed by a HRP-conjugated goat anti-mouse IgG antibody in TBST containing 2% nonfat dried milk. Bands were visualized using SuperSignal West Pico or Femto Maximum Sensitivity Substrate™ and exposure in a Fuji Film LAS 3000. Blots were then stripped using Restore™ Western blot stripping reagent and probed for the presence of actin using an anti-actin antibody followed by the HRP-conjugated goat anti-mouse IgG and visualization as described. Intensity of acquired bands was determined using Multigauge software.

Analysis of sphingosine kinase 1 (SphK1) activity: Cell lysate SphK1 activity was analyzed by monitoring the amount of sphingosine 1 phosphate (S1P) synthesized within a specified time as previously described (30). In brief, after cells were washed with phosphate buffered saline, they were lysed by incubating in lysis buffer (20 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 10 mM NaF, 1 mM

Na3VO4) containing protease inhibitors (Complete Mini EDTA-Free Protease inhibitor tablet, Cat. #11 836 170 001, Roche) for 10 min. at 4˚C. After centrifugation (100,000 x g, 15 min) to remove insoluble material, protein content of each lysate was determined using the BCA protein assay (Pierce). Lysates prepared as described contained both cytosolic and membrane-associated SphK1 (31). Lysate samples containing 25µg of protein were combined with 25 µM D-erythro-sphingosine, 10 mM MgCl2, 2 µM ATP and 2 µCi [-32P] ATP in a 100 µL final reaction volume of SKAAB buffer (20 mM Tris

156 pH 7.4, 1 mM DTT, 1 mM EDTA, 0.1% Triton X-100, 1 mM Na3VO4, 15 mM NaF, 0.5 mM 4-deoxypyridoxine) and incubated, with shaking, at 37˚C for 10 min. Kinase activity was inhibited by addition of 10µl of 6N HCl and labeled lipids extracted by first adding

400µl of CHCl3:CH3OH (1:2, v:v) followed by addition of 125µl CHCl3 and 125µl 1M

KCl to enhance phase separation. After drying the chloroform phase under N2, recovered lipids were taken up in 20µl of CHCl3 and analyzed by TLC. Silica gel coated plates (Whatman, Florham Park, NJ) were developed in butanol:water:acetic acid (3:1:1, v:v:v) and bands visualized by exposure to x-ray film. Label associated with bands moving with the same Rf as S1P (0.32) was determined.

Results

To determine whether the H63D- or C282Y-mutations had significant effects on cell growth and morphology, growth curves were determined and cell morphology visualized initially by phase contrast microscopy. Morphologically, cells expressing C282Y HFE did not appear to be as elongated nor to develop the processes seen on those stably transfected to produce WT-HFE (Figure 1A). While cells expressing H63D-HFE appeared to be more similar to those expressing WT-HFE, they grew more slowly and were not as elongated (Figure 1B).

The presence/absence of WT-HFE in lipid rafts was ascertained because it was shown to associate with TfR2, a lipid raft component. Upon isolation of lipid rafts from SH-SY5Y cells transfected with WT-HFE, the HFE was found in the same fractions as flotillin (Figure 2), a lipid raft marker (32).

Because ganglioside content has been implicated in cell migration and differentiation (eg. 33,34, respectively), ganglioside analyses were carried out. The results indicated that the ganglioside composition differed between cells expressing WT-HFE and those expressing C282Y-HFE (Figure 3). Those expressing the C282Y-HFE variant expressed

157 a resorcinol positive band that migrated close to GM3 while the major resorcinol positive band in both the WT- and H63D-HFE transfected cells appeared to migrate as GM2. Interestingly, cells expressing WT-HFE appeared to express more of a resorcinol positive component that migrated between gangliosides GD1a and GD1b than was present in cells expressing either the H63D- or C282Y-HFE mutant.

Despite the observation that CTxB can bind to an unidentified protein (35) in addition to gangliosides, the changes in ganglioside composition led us to look at binding of Alexa- Fluor®594-CTxB to the surface of cells transfected with either WT-, H63D-, or C282Y- HFE. Essentially no adherence was seen to the surface of cells expressing C282Y-HFE compared with that to cells expressing WT- or H63D-HFE . Interestingly, the latter appeared to have more CTxB adhering to cell processes than cells expressing WT-HFE (Figure 4).

In light of the observed changes in ganglioside composition, RT-qPCR analyses were done to determine possible effects of expression of a mutant HFE on sphingolipid metabolism. The results indicated that transcription of specific genes involved in the synthesis of sphingolipids differed in cell lines expressing HFE variants. Five of the 14 genes studied (see Figure 5 for reactions catalyzed) were over-expressed (SPHK1, SGPP1, GalT1, ST3Gal1, and ST8Sia5), in cells transfected with C282Y-HFE and 2 were under expressed (ST8Sia1 and ß4GalNT) by a factor of 2 or more. In contrast, cells transfected with H63D-HFE over expressed one gene (ST8Sia5) and under expressed two (ST8Sia1 and SPHK1, Table 1). The reduction in transcription of both GD3 and GM2 synthase is consistent with the ganglioside pattern obtained for cells transfected with the C282Y-HFE variant, while the decrease in GD3 synthase found for cells expressing H63D-HFE may account for the relative increase in expression of the band moving as GM2 relative to that moving between GD1a and GD1b (Figure 3).

158 Western blot analyses were done to ascertain whether the increase in transcription of SphK1 resulted in increased expression of the enzyme relative to actin, one of the two controls used to determine changes in transcription. It can be seen that the ratio of SphK1/actin was significantly greater in the cells expressing C282Y-HFE than in cells expressing either WT- or H63D-HFE (Figure 6). However, it also appeared as though the expression of actin might be somewhat less in the C282Y-HFE-expressing cells than in those expressing WT-HFE. To address this possibility, cell lysate synthesis of S1P was monitored. It can be seen (Figure 7) that significantly more S1P was synthesized by cell lysates from C282Y-HFE-expressing cells than by those expressing either WT- or H63D- HFE.

Discussion

A variety of functions have been identified for sphingolipids. For example, sphingosine and ceramide, have been implicated as initiators of apoptosis and S1P as a promoter of cell survival and angiogenesis (36) while glycosphingolipids have been shown to function in signal transduction (eg. 37), cell proliferation and differentiation (eg. 38), and Ca2+ transport (39). The association of sphingolipids with lipid rafts allows them to “fine tune” signal transduction (17) mediated by raft-associated proteins and it is well known that changes in ganglioside composition are associated with changes in neuronal development and differentiation (40). Therefore, it is not surprising that mutations in WT-HFE that perturb its association with lipid rafts might over time induce changes in the composition of rafts that in turn affect cell behavior in ways other than by altering iron uptake. For example, elevated levels of GM3 have been shown to disrupt interaction of the insulin receptor with caveolin-1 causing the cell to become insulin resistant (41), and a number of reports have described the effects of different gangliosides on various growth factor receptors that undergo autophosphorylation when bound by the appropriate ligand (42-44). For neurons, such changes might contribute to the effect of the H63D mutation on the age of onset of AD (2), or to the putative protective role of the C282Y mutation in AD (5).

159 Results of the studies presented here indicate that while changes induced by expression of the H63D variant of HFE appeared to be less dramatic than those induced by expression of C282Y-HFE, it did induce significant (>2-fold) decreases in transcription of SphK1 and GD3 synthase, and an increase in transcription of ST8SIA5. In terms of ganglioside expression, a marked difference was seen between the relative amount of ganglioside with the same mobility as GM2 and one with mobility between GD1a and GD1b, possibly GT1a. Changes in morphology were also noted with cells transfected with H63D-HFE proliferating more slowly and appearing to express more processes bound by Alexa-Fluor®-594 conjugated CTxB. The differences in CTxB binding to the surface of cells expressing the different variants may reflect differences in ganglioside expression. Surface Plasmon resonance analyses of the binding of cholera toxin to specific gangliosides indicated that it adhered best to GM1 followed by GM2>GD1a>GM3 (45). Our analysis of the ganglioside content of C282Y-HFE-expressing cells showed that they expressed a ganglioside that migrated similarly to GM3 while WT- and H63D-HFE- expressing cells had more of a ganglioside that migrated like GM2, with cells expressing the H63D variant appearing to express relatively more GM2 than those expressing WT- HFE. The relative increase in GM2 could account for the increase in CTxB adherence to cells expressing the H63D variant while the preponderance of a ganglioside having a mobility similar to that of GM3 could account for the negligible adherence of CTxB to cells expressing C282Y-HFE.

In addition to its effect on ganglioside composition, expression of C282Y-HFE in SH- SY5Y cells resulted in altered morphology and a significant increase in transcription and expression of SphK1. The increase in expression of SphK1 was accompanied by an increase in the amount of the pro-survival lipid, S1P synthesized by cell lysates. High expression of SphK1 activity in breast cancers and glioblastomas was found to correlate with poor survival (46,47), an observation that supports the role of cellular S1P as a bioactive “ pro-survival” lipid (48). The increase in S1P may contribute to the enhanced drug resistance conferred by the induction of p16INK4A, previously observed in studies of C282Y-HFE-transfected SH-SY5Y cells (49). Both findings may help to explain why expression of the C282Y-HFE variant is a risk factor for colon and breast cancers (4).

160 These findings also provide support for studying the efficacy of small molecule inhibitors of SphK1 as potential inhibitors of cancer development (36,50).

While much focus has been placed on the association of expression of C282Y-HFE and cancer, it was recently shown that there was a significant negative association of this mutation with AD suggesting that it might have a protective role in neurodegeneration (5). This could be due to the C282Y-induced changes in sphingolipid composition. Potential effects of those changes include: 1) increased expression of S1P contributing to cell survival, 2) altered ganglioside composition reducing the probability of assembly of amyloid ß-protein variants, 3) lipid changes altering signal transduction, 4) their altering Ca2+ transport, or 5) a combination of small changes that over time enhance cell survival. Development of an inhibitor(s) of SphK1 as a treatment for colon or breast cancer does not mean that the potential protective effect of the C282Y mutation on AD would be nullified provided the drug did not cross the blood brain barrier. In the case of glioblastoma multiforme where up-regulation of SphK1 has been linked to poor prognosis (51), the use of nanovesicles to both transport such an inhibitor across the BBB and to target it specifically to tumor cells as previously described (52) might provide an effective therapy.

In conclusion, the results described indicate that both the H63D and C282Y variants of HFE induced significant changes in the sphingolipid composition of the SH-SY5Y human neuroblastoma cells in which they were expressed. Knowledge of these differences coupled with the effects expression of these variants have been found to have in people may provide insights useful for development of potential therapies.

Acknowledgements

The authors thank Dr. Jong Yun for helpful discussions, Dr. K.A. Petro for her help with isolating lipid rafts, and Dr. Sang Lee for providing us with the SH-SY5Y cells stably transfected to express either FLAG-tagged WT-, H63D- or C282Y-HFE alleles. We also thank the Jake Gittlen Cancer Research Foundation (JAH), and the George M. Leader Family Corporation (JRC) for their support.

161 Table 3.1: Relative expression of mRNAs for 14 enzymes involved in sphingolipid synthesis

162 Figure 3.1: Cell morphology and growth curves. Phase contrast photomicrographs for SH-SY5Y human neuroblastoma cells stably transfected with either wild type HFE (WT), H63D HFE (H63D), or C282Y HFE (C282Y). Cells were seeded at 6X105 per well on a six well plate and photographed after 5.5 days in culture. The bar on each photomicrograph indicates 140 microns. Cells were allowed to grow for 72 hr prior to taking cell counts at 12hr intervals. Green indicates growth of cells transfected with WT- HFE; blue, C282Y-HFE; and red H63D-HFE.

163

Figure 3.2: Association of FLAG-tagged WT-HFE with lipid rafts. Blots were probed with a mouse monoclonal anti-flotillin antibody (upper bands) or an anti-FLAG antibody (lower bands) prior to exposure to a secondary HRP-conjugated goat anti-mouse antibody. Numbers refer to 1 ml fractions taken from the top (1) to the bottom of the gradient (10). Bands were visualized using SuperSignal West Femto Maximum Sensitivity Substrate™ and exposure in and exposure in a Fuji Film LAS 3000. Location of flotillin visualized on the same samples as the HFE or variant thereof is shown to account for any variation in the individual gradients or the possible effect of HFE or variant thereof on its location.

164 Figure 3.3: HPTLC analysis of the ganglioside composition of cells transfected with either WT, H63D, or C282Y variants of HFE. Isolated gangliosides were separated by HPTLC on plates developed in either A) chloroform:isopropanol:50mM KCl (2:13.4:4.6, v:v:v); or B) CHCl3:CH3OH:0.3% CaCl2•2H2O (60:30:8, v:v:v).

165 Figure 3.4: Binding of Alexa-Fluor®594-conjugated CTxB to the cell surface of SH- SY5Y human neuroblastoma cells stably transfected with either WT-, H63D-, or C282Y-HFE. Nuclei were visualized using DAPI; bars on the photomicrographs indicate 20 microns.

166 Figure 3.5: Steps in sphingolipid metabolism catalyzed by the enzymes studied. , sialic acid; , glucose; , galactose; and , N-acetylgalactosamine. While linkages are not indicated, the gangliotetaose carbohydrate sequence is Galß1-3GalNAcß1- 4Galß104Glcß1-. See Table 1 for definitions of gene abbreviations. Nomenclature used is that of Svennerholm (53).

167 Figure 3.6: Expression of sphingosine kinase 1. A) Western blot analysis of the expression of SPhK1 and actin, B) ratio of SphK1/actin (n=3). ** indicates that the difference between the ratio for C282Y-HFE expressing cells and those expressing WT- or H63D-HFE was significant (p<0.001; one way analysis of variance followed by the Newman-Keuls multiple comparison test).

168 Figure 3.7: S1P synthesized by cell lysate SphK1. A) 32P-labeled lipid extracts were separated by TLC using butanol:water:acetic acid (3:1:1, v:v:v) as the running solvent and bands visualized by exposure to x-ray film. B) Label associated with S1P synthesized by cell lysates from cells expressing C282Y-HFE was significantly greater [*p<0.01 (n=3), one way analysis of variance followed by the Newman-Keuls multiple comparison test] than that for cells expressing WT- or H63D-HFE.

A B

169 References

1. Goswami T, Andrews NC (2006) Hereditary hemochromatosis protein, HFE, interaction with transferrin receptor 2 suggests a molecular mechanism for mammalian iron sensing. J. Biol. Chem. 281:28494-28498 2. Combarros O, Garcia-Roman M, Fontalba A et al (2003) Interaction of the H63D mutation in the hemochromatosis gene with the apolipoprotein E epsilon 4 allele modulates age at onset of Alzheimer's disease. Dement. Geriatr. Cogn. Disord. 15:151-154 3. Connor JR, Lee SY (2006) HFE mutations and Alzheimer’s disease. J. Alzheimers Dis. 10:2267-276 4. Osborne NJ, Gurrin LC, Allen KJ et al (2010) HFE C282Y homozygotes are at increased risk of breast and colorectal cancer. Hepatology 51:1311-1318 5. Correia AP, Pinto JP, Dias V, et al (2009) CAT53 and HFE alleles in Alzheimer’s disease: a putative protective role of the C282Y HFE mutation. Neurosci. Lett. 457:129-132 6. Waheed A, Parkkila S, Zhou XY, et al (1997) Hereditary hemochromatosis: effects of C282Y and H63D mutations on association with beta2-microglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proc. Natl. Acad. Sci. USA 94:12384-12389 7. Lawless MW, Mankan AK White M et al (2007) Expression of hereditary hemochromatosis C282Y HFE protein in HEK293 cells activates specific endoplasmic reticulum stress responses. BMC Cell Biol. 8:30 8. Feder JN, Penny DM, Irrinki A et al (1998) The hemochromatosis gene product complexes with the transferrin receptor and lowers its affinity for ligand binding. Proc. Natl. Acad. Sci. USA 95:1472-1477 9. Calzolari A, Raggi C, Deaglio S et al (2006) TfR2 localizes in lipid raft domains and is released in exosomes to activate signal transduction along the MAPK pathway. J. Cell Sci. 119:4486-4498 10. Chen J, Chloupkova M, Gao J, et al (2007) HFE modulates transferring receptor 2 levels in hepatoma cells via interations that differ from transferrin receptor 1-HFE interactions. J. Biol. Chem. 282:36863-36870

170 11. Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387:569-572 12. Sakiyama H, Robbins PW (1973) Glycolipid synthesis and tumorigenicity of clones isolated from the Nil 2 line of hamster embryo fibroblasts. Fed. Proc. 32:86-90 13. Brady RO, Fishman PH, Mora PT (1973) Membrane components and enzymes in virally transformed cells. Fed. Proc. 32:102-108 14. Hakomori S (1981) Glycosphingolipids in cellular interaction, differentiation, and oncogenesis. Ann. Rev. Biochem. 50:733-764 15. Petro KA, Schengrund C-L (2009) Membrane raft disruption promotes axonogenesis in N2a neuroblastoma cells. Neurochem. Res. 34:29-37 16. Ruckhaberle E, Karn T, Rody A et al (2009) Gene expression of ceramide kinase, galactosyl ceramide synthase and ganglioside GD3 synthase is associated with prognosis in breast cancer. J. Cancer Res. Clin. Oncol. 135:1005-1013 17. Fantini J, Barrantes FJ (2009) Sphingolipid/cholesterol regulation of neurotransmitter receptor conformation and function. Biochim. Biophys. Acta 1788:2345-2361 18. Miyagi T, Wada T, Yamaguchi K, et al (2008) Plasma membrane-associated sialidase as a crucial regulator of transmembrane signalling. J Biochem 144:279-285 19. Allende ML, Proia RL (2002) Lubricating cell signaling pathways with gangliosides. Curr Opin Struct Biol 12:587-592 20. Yamamoto N, Hirabayashi Y, Amari M, et al (2005) Assembly of hereditary amyloid beta-protein variants in the presence of favorable gangliosides. FEBS Lett. 579: 2185-2190 21. Yamamoto N, Matsuzaki K, Yanagisawa K (2005) Cross-seeding of wild-type and hereditary variant-type amyloid beta-proteins in the presence of gangliosides. J. Neurochem. 95: 1167-1176 22. Yanagisawa M, Ariga T, Yu RK (2010) Cytotoxic effects of G(M1) ganglioside and amyloid beta-peptide on mouse embryonic neural stem cells. ASN Neuro. 2: e00029 23. Tamboli (2005) Inhibition of glycosphingolipid biosynthesis reduces secretion of the ß-amyloid precursor protein and amyloid ß-peptide. J. Biol. Chem. 280:28110-28117

171 24. Cecchi C, Nichino D, Zampagni M, et al (2009) A protective role for lipid raft cholesterol against amyloid-induced membrane damage in human neuroblastoma cells. Biochim. Biophys. Acta 1788:2204-2216 25. Wang XS, Lee S, Simmons Z, et al (2004) Increased incidence of the HFE mutation in amyotropic lateral sclerosis and related cellular consequences. J. Neural. Sci. 227:27-33 26. Lee SY, Patton SM, Henderson RJ, et al (2007) Consequences of expressing mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking endogenous HFE. The FASEB J. 21:564-576 27. Petro KA, Dyer MA, Yowler, BC, et al (2006) Disruption of lipid rafts enhances activity of botulinum neurotoxin serotype A. Toxicon 48:1035-1045 28. Kasperzyk JL, d’Azzo A, Platt FM, et al (2005) Substrate reduction reduces gangliosides in postnatal cerebrum-brainstem and cerebellum in GM1 gangliosidosis mice. J. Lipid Res. 46:744-751 29. Svennerholm L (1957) Quantitative estimation of sialic acids. II. A colorimetric resorcinol-hydrochloric acid method. Biochim. Biophys. Acta 24:604-611 30. Hengst JA, Guilford JM, Conroy EJ, et al (2010) Enhancement of sphingosine kinase 1 catalytic activity by deletion of 21 amino acids from the COOH-terminus. Arch. Biochem. Biophys. 494:23-31 31. Hengst JA, Guilford JM, Fox TE, et al (2009) Sphingosine kinase 1 localized to the plasma membrane lipid raft microdomain overcomes serum deprivation induced growth inhibition. Arch. Biochem. Biophys. 492:62-73 32. Bickel PE, Scherer PE, Schnitzer JE, et al (1997) Flotillin and epidermal surface antigen define a new family of caveolae-associated integral membrane proteins. J. Biol. Chem. 272:13793-13802 33. Johnstone SR, Stallcup WB (1988) Altered expression of D1.1 ganglioside in the cerebellum of the Weaver mouse. J. Neurochem. 51:1655-1657 34. Purpura DP, Baker HJ (1978) Meganeurites and other aberrant processes of neurons in feline GM1-gangliosidosis: a Golgi study. Brain Res. 116:1-21

172 35. Yanagisawa M, Ariga T, Yu RK (2006) Cholera toxin B subunit binding does not correlate with GM1 expression: a study using mouse embryonic neural precursor cells. Glycobiol. 16: 19G-22G 36. Cuvillier O, Ader I, Bouquerel P, et al (2010) Activation of sphingosine kinase-1 in cancer: implications for therapeutic targeting. Curr. Mol. Pharmacol. 3:53-65 37. Duchemin AM, Ren Q, Neff NH, et al (2008) GM1-induced activation of phosphatidylinositol 3-kinase: involvement of Trk receptors. J. Neurochem. 104:1466-1477 38. Kwak DH, Yu k, Kim SM, et al (2006) Dynamic changes of gangliosides expression during the differentiation of embryonic and mesenchymal stem cells into neural cells. Exp. Mol. Med. 38:668-676. 39. Wu G, Xie X, Lu ZH, et al (2009) Sodiium-calcium exchanger complexed with GM1 ganglioside in nuclear membrane transfers calcium from nucleoplasm to endoplasmic reticulum. Proc. Natl. Acad. Sci. USA 106:10829-10834 40. Rosner H (1982) Ganglioside changes in the chicken optic lobes as biochemical indicators of brain development and maturation. Brain Res. 236:49-61 41. Kabayama K, Sato T, Saito K, et al (2007) Dissociationof the insulin receptor and caveolin-1 complex by ganglioside GM3 in the state of insulin resistance. Proc. Natl. Acad. Sci. USA 104:13678-13683 42. Van Brocklyn J, Bremer EG, Yates AJ (1993) Gangliosides inhibit platelet-derived growth factor-stimulated receptor dimerization in human glioma U-1242MG and Swiss 3T3 cells. J. Neurochem. 61:371-374 43. Mutoh T, Tokuda A, Miyadai T (1995) Ganglioside GM1 binds to the Trk protein and regulates receptor function. Proc. Natl. Acad. Sci. USA 92:5087-5091 44. Rusnati M, Urbinati C, Tanghetti E, et al (2002) Cell membrane ganglioside GM1 ganglioside is a functional coreceptor for fibroblast growth factor 2. Proc. Natl. Acad. Sci. USA 99:4367-4372 45. Kuziemko GM, Stroh M, Stevens RC (1996) Cholera toxin binding affinity and specificity for gangliosides determined by surface Plasmon resonance. Biochemistry 35:6375-6384

173 46. Van Brocklyn JR, Jackson CA, Pearl DK et al (2005) Sphingosine kinase-1 expression correlates with poor survival of patients with glioblastoma multiforme: roles of sphingosine kinase isoforms in growth of glioblastoma cell lines. J. Neuropathol. Exp. Neurol. 64:695-705 47. Ruckhaberle E, Rody A, Engels K et al (2008) Microarray analysis of altered sphingolipid metabolism reveals prognostic significance of sphingosine kinase 1 in breast cancer. Breast Cancer Res. Treat. 112:41-52 48. Brizuela L, Dayon A, Doumerc N, et al (2010) The sphingosine kinase-1 survival pathway is a molecular target for the tumor-suppressive tea and wine polyphenols in prostate cancer. FASEB J EPub: PMID: 20522783 49. Lee SY, Liu S, Mitchell RM, et al (in press) HFE polymorphisms influence the response to chemotherapeutic agents via induction of p16INK4A. Int. J. Cancer 50. Chumanevich AA, Poudyal D, Cui X, et al (2010) Suppression of colitis-driven colon cancer in mice by a novel small molecule inhibitor of sphingosine kinase. Carcinogenesis Epub PMID: 20688834 51. Kapitonov D, Allegood JC, Mitchell C, et al (2009) Targeting sphingosine kinase 1 inhibits Akt signaling, induces apoptosis, and suppresses growth of human glioblastoma cells and xenografts. Cancer Res. 69:6915-6923. 52. Madhankumar AB, Slagle-Webb B, Wang X, et al (2009) Efficacy of IL-13 receptor targeted liposomal doxorubicin in the intracranial brain tumor model. Molec. Cancer Therapeutics 8:648-654 53. Svennerholm, L (1980) Ganglioside designation. Adv. Exp. Med. 125: 11

174 Chapter 4

C282Y increases cellular proliferation by inducing cholesterol accumulation

Abstract

Disruptions in the maintenance of iron and cholesterol metabolism have also been implicated in several cancers. Therefore, the effects of C282Y-HFE on lipid (cholesterol and sphingolipid) metabolism and its possible relevance to cancer phenotype was also studied. Expression of C282Y-HFE altered transcription of a number of enzymes involved in cholesterol and sphingolipid metabolism. More specifically it was associated with significant changes in the expression of genes involved in cholesterol metabolism were consistent with significant elevation in total cholesterol content of C282Y-HFE cells. Of particular interest was the finding that expression of the enzyme sphingosine kinase 1(SphK) and its pro-survival metabolite, sphingosine-1-phosphate (S1P) was elevated. These effects may contribute to the increased risk of colorectal, breast, prostate, ovarian, and brain tumors reported for carriers of C282Y-HFE and explain why it is considered protective against Alzheimer’s disease while H63D-HFE is a risk factor for neurodegenerative diseases. Changes in the gene expression seen for proteins involved in cholesterol metabolism, indicate that increased uptake of cholesterol is the likely cause for the elevation in cellular cholesterol content seen in C282Y-HFE expressing cells. Since simvastatin is an inhibitor of cholesterol synthesis and has been proposed to be beneficial for the treatment of certain cancers, its effect on the viability of C282Y-HFE cells was ascertained. A higher concentration of simvastatin was needed to decrease survival of cells expressing C282Y-HFE than WT-HFE. Because of the significant increase of SphK in C282Y cells, the effect of a SphK inhibitor (SKI) on cell survival was monitored. Again, WT-HFE cells were more sensitive to this treatment than C282Y- HFE cells, reflecting the drug resistant phenotype of C282Y-HFE cells. Treatment of cells with both simvastatin and SKI revealed a small effect on survival. Collectively,

175 these studies indicated that C282Y-HFE markedly affected both cholesterol and sphingolipid metabolism and support the possibility that inhibition of both cholesterol synthesis/uptake and as well as synthesis of S1P as a potential therapeutic strategy for treatment of cancer in carriers of C282Y-HFE.

Introduction

Iron and cholesterol are both essential for normal cellular growth and proliferation. They are also needed for tumor growth and metastasis. In this chapter, the role of iron and cholesterol in carcinogenesis will be described and data are presented to show the effects of the C282Y variant of HFE and resulting iron-overload on cholesterol metabolism. Finally, the relationship between these pathways and its relevance to cancer will be discussed.

Role of iron in cancer:

Iron is a required co-factor for a number of key cellular enzymes involved in cellular proliferation including DNA, protein and lipid synthesis as well as for oxygen transport as a component of heme. Heme is a major component of hemoglobin (red pigment in blood) and several enzymes. Most of the iron in the body is found in heme, or in ferritin (Ft), an iron storage protein, or bound to transferrin (Tf), an iron transport protein. The ability of one molecule of ferritin to bind up to 4500 molecules of iron makes it the major source of iron (Stojiljkovic and Perkins-Balding, 2002). Tf that has two binding sites for iron is primarily involved in transporting iron in the circulation. Body iron status is usually measured indirectly in terms of dietary iron intake, iron supplement intake, serum ferritin levels, serum iron (also called Tf iron), and transferrin saturation (total iron binding capacity of Tf).

Evidence for the role of iron in cancer was first reported when intramuscular injection of iron induced malignant tumors in rats (Richmond, 1959). Moreover, studies by Ba et al showed that when iron chelation was used to decrease serum iron concentration in athymic mice, tumor growth was suppressed (Ba et al., 2011), indicating a need for high

176 iron levels for cancer growth. Consistent with these experimental findings, patients that were treated with iron preparations were shown to develop sarcomas (Greenberg, 1976). In addition, Zacharski et al. reported reduction in the recurrence of cancer or cancer- related deaths in individuals undergoing frequent phlebotomy (Zacharski et al., 2008). These observations led to several epidemiological studies, both prospective and retro- prospective, the results of which showed increased risk of cancer in patients with exposure to iron treatment (Graf and Eaton, 1985; Kazantzis, 1981; Selby and Friedman, 1988; Stevens et al., 1986; Stevens et al., 1988). In particular, association of iron with increased risk of colorectal cancer has been repeatedly identified in epidemiological studies (Nelson, 2001; Stevens et al., 1988; Wurzelmann et al., 1996). A meta-analysis of thirty-three studies revealed an association of iron with increased risk of colorectal neoplasia in 75% of the studies analyzed (Nelson, 2001).

At the mechanistic level, it has been proposed that the presence of excessive iron can be specifically detrimental to the cell because iron can catalyze production of oxidative radicals which can cause lipid peroxidation, protein and DNA damage that results in mutagenesis. In fact, accumulation of oxidative radicals from unabsorbed dietary iron in the colon has been proposed to be the likely cause for the higher incidence of colorectal cancer (Lund et al., 1999; Lund et al., 1998). This idea was supported by the findings of increased proliferation of colonic epithelium as measured in vivo by measuring incorporation of 3H-thymidine into colonic mucosa of rats fed with heme (Sesink et al., 1999). Furthermore, animal studies utilizing HFE-knockout mice fed with a standard iron diet showed increased concentration of malondialdehyde (MDA; indicator of lipid peroxidation) in colon tissue relative to animals fed a low iron diet, indicating a synergism between dietary iron content and HFE genotype and increasing oxidative damage to colon tissue (Stevens et al., 2003). In addition to colorectal cancer, iron accumulation is also associated with liver cancer. Compared to samples from normal livers, liver samples from patients with hemochromatosis (an iron-overload disorder) had significantly higher levels of iron, and showed development of fibrosis, cirrhosis, and hepatoma (Bassett et al., 1986). Epidemiological studies have shown that in European countries, where the incidence of hemochromatosis is much higher, individuals with hemochromatosis were 23 times more likely to develop hepatocellular carcinoma (HHC)

177 than individuals without it (Lauret et al., 2002; Milman et al., 2001; Niederau et al., 1996). The mechanism by which iron accumulation may be a risk factor for cancer is by induction of oncogenes such as MDM2 (murine double minute 2), which in turn is a regulator of p53 (Dongiovanni et al., 2010). Iron-induced activation of p53 activity via MDM2 has been shown to increase risk of liver cirrhosis and HHC in humans (Dongiovanni et al., 2010), as well as, in a rat model of hepatocarcinogenesis (Mizukami et al., 2010).

In addition to free iron, elevated body iron stores are also implicated in the risk of cancer. High amounts of ferritin in the serum are associated with metastasis of hepatocellular carcinoma (Hellerbrand et al., 2003), renal cell carcinoma (Miyata et al., 2001; Wu et al., 2004), and cerebrospinal fluid ferritin levels are increased in glioblastoma patients (Sato et al., 1998). Therefore, studies of genetic mutations resulting in iron-overload such as that induced by HFE variants should be useful in developing cancer therapeutics.

C282Y-HFE in cancer

Cellular iron uptake is regulated by the hemochromatosis gene product HFE. The wild type (WT)-HFE protein interacts with the transferrin receptor to limit uptake of iron- loaded transferrin into cell. In contrast to WT-HFE, the C282Y-HFE variant is retained in the endoplasmic reticulum (ER) leading to increased iron uptake (Feder et al., 1997; Waheed et al., 1997). A number of studies have shown that C282Y-HFE is associated with accumulation of iron at the cellular and tissue level (Camaschella et al., 2002; Lee et al., 2007a). C282Y-HFE cells express less TfR; this is an indicator of iron-overload as the cell tries to limit further iron uptake by reducing TfR expression (Lee et al., 2007a).

178 Because C282Y-HFE is the prime suspect in the iron-overload disorder hereditary hemochromatosis (HH), one of the complications of which is hepatocellular carcinoma, and iron accumulation has been associated with malignancy, assessment of C282Y-HFE as a risk factor for different types of cancer is of interest. In fact, carriers of C282Y-HFE have been found to have an increased risk of breast, ovarian, colorectal, and prostate cancer relative to WT-HFE carriers (Gannon et al., 2011; Osborne et al., 2010; Shaheen et al., 2003; Syrjakoski et al., 2006). Considering that the prevalence of homozygosity for the C282Y mutation varies between 3% and 12% in Caucasians (Fargion et al., 2010) and the lack of knowledge about its prevalence in different types of cancers, a large number of people may be at an increased risk of cancer. The mechanism whereby C282Y-HFE influences the incidence of cancer is unknown.

Though it is unclear how C282Y-HFE can contribute to cancer, evidence from cell culture studies is accumulating to show effects on key cellular mechanisms. For example, expression of C282Y-HFE in human neuroblastoma cells resulted in an elevation in intracellular iron, increased lipid peroxidation, and increased cell proliferation (Lee et al., 2007b). In addition, cells expressing C282Y-HFE are resistant to induction of differentiation by retinoic acid and to several chemotherapeutic drugs i.e. doxorubicin etc as a result of induction of expression of the cyclin-dependent kinase inhibitor, p16INK4a (Lee et al., 2010). C282Y-HFE expressing cells also exhibit significantly higher expression of sphingosine kinase 1 (SphK1), its pro-survival metabolite sphingosine 1 phosphate (S-1-P), and increased cellular proliferation (Ali-Rahmani et al., 2011). In line with the cell culture findings, a recent epidemiological study reported that carriers of the C282Y-HFE variant had alterations in their serum lipid profile (Adams et al., 2009; Pankow et al., 2008). Though, these studies suggest a link between iron and lipid metabolism, current knowledge in this area is very limited. So far, no studies have been conducted to assess effects of C282Y-HFE on cholesterol metabolism directly. Therefore, studies were conducted to explore this connection in this thesis. Independent of C282Y- HFE, the role of cholesterol and other lipids has been investigated and a brief overview is presented below.

179

Role of cholesterol in cancer:

The high proliferation of cancer cells requires large amounts of lipids as building blocks for biological membranes. Accumulation of cholesterol has been reported to occur in various malignancies, including prostate cancer and oral cancer (Freeman and Solomon, 2004; Kolanjiappan et al., 2003; Zhuang et al., 2003; Zhuang et al., 2002), myeloid leukemia, lung, and breast cancers (Bennis et al., 1993; Duncan et al., 2004; El-Sohemy and Archer, 2000; Li et al., 2003). Li et al. showed higher levels of membrane cholesterol and lipid rafts in breast and prostate tumor cells compared with normal cells from the same tissues (Li et al., 2006). In addition, several tumors were reported to exhibit increased expression of HMGCoAR and mevalonate (cholesterol precursor) (Duncan et al., 2004). Administration of mevalonate to nude mice with a mammary tumor xenograft promoted tumor growth (Duncan et al., 2004), further supporting the link between elevated cholesterol content and cancer.

A strong link between cholesterol and cancer is also supported by findings that elevated serum cholesterol increased expression of cholesterol synthesis enzymes in the cancer cells. The mechanism whereby cholesterol influences cancer growth and/ migration remains unclear. Recent evidence points at regulation of AKT dependence on cholesterol and role of lipid rafts Several studies have shown that changes in membrane lipid composition, either through the generation of sphingosine-1-phosphate or an increase in membrane cholesterol alters signaling pathways including AKT mediated cell survival (Oh et al., 2007; Osawa et al., 2005; Zhuang et al., 2002). Akt/protein kinase B (PKB) is a serine/threonine kinase that is a critical regulator of cell survival and proliferation, especially in human malignant cancer. Phosphorylation of AKT results in its activation, which in turn phosphorylates and inactivates its downstream effecters such as pro- apoptotic proteins. Interestingly, activated AKT can, in turn, affects cholesterol metabolism by inducing synthesis of the sterol regulatory element binding protein and transcription of genes such as HMGCoA synthase and HMGCoAR (Freeman and Solomon, 2004; Porstmann et al., 2005). Akt activation also up-regulates anti-apoptotic genes such as Bcl-xL and FLICE-inhibitory protein (FLIP) (Cardone et al., 1998; Datta et

180 al., 1997; Kane et al., 1999; Panka et al., 2001; Romashkova and Makarov, 1999; Shimamura et al., 2003). A subpopulation of AKT resides in lipid rafts and mediates cell survival signaling (Adam et al., 2007; Zhuang et al., 2001). Interestingly, Akt was shown to induce cholesterol synthesis by inducing sterol-regulatory element binding protein and transcription of multiple genes involved in fatty acid and cholesterol metabolism, including HMG-CoAR, HMG-CoA synthase, and fatty acid synthase (Porstmann et al., 2005).Taken together, these studies have shown that AKT activation is dependent upon cholesterol content of lipid rafts suggesting the existence of a potential feedback mechanism, whereby cholesterol-induced recruitment of AKT in lipid rafts and subsequent activation may result in enhanced cholesterol synthesis, increase in lipid rafts and increased localization of AKT into rafts (Adam et al., 2007) . These findings indicate a direct relationship between cholesterol, AKT activation, and cell survival (Adam et al., 2007; Elhyany et al., 2004; Hill et al., 2002; Oh et al., 2007; Partovian and Simons, 2004; Zhuang et al., 2001).

In the previous chapter, we have shown that WT-HFE is localized in lipid rafts, membrane microdomains in the plasma membrane that are enriched in cholesterol and sphingolipids. Lipid rafts provide a platform for protein-protein interactions and play an important role in signal transduction mediating cell proliferation etc. We have shown that C282Y-HFE, unlike WT-HFE, is not associated with lipid rafts (Ali-Rahmani et al., 2011), indicating that its interaction with other raft-resident proteins could be disrupted. Based on our previous finding that sphingolipid metabolism was altered in cells expressing C282Y-HFE, we investigated the effect of its expression on cholesterol, another lipid enriched in lipid rafts. We hypothesized that the expression of C282Y-HFE would result in disruption of cholesterol metabolism. To test this hypothesis, we determined the total cholesterol content and expression of genes involved in cholesterol metabolism in C282Y-HFE-expressing human neuroblastoma cells. We also tested whether inhibition of cholesterol and/ or sphingosine kinase affected survival of C282Y- HFE cells.

181 Methods and materials

Cell culture:

SH-SY5Y human neuroblastoma cells, stably transfected (Lee et al., 2007b) to express either FLAG-tagged wild-type (WT)- or C282Y-HFE alleles, were grown in DMEM/F12 medium containing 10% fetal bovine serum, 1% antibiotics (Penn-Strep, glutamine), 1% non-essential amino acids, and 1.8g/L NaHCO3 at 37˚C in an atmosphere of 5%

CO2/95% air.

Cholesterol analyses:

Lipids were extracted from equal number of cells (106) of each cell type using chloroform: isopropanol:Triton X-100 (7:11:0.1). The CHCl3 (bottom) layer was collected, dried, and extracted material resuspended in cholesterol reaction buffer provided with the Biovision kit. Cholesterol content was measured using the cholesterol quantification kit according to the manufacturer’s (Biovision) directions. After determining protein content using the Dc Protein assay (BioRad), cholesterol content was normalized to protein content.

Gene expression:

Total RNA was isolated from SH-SY5Y cells expressing either WT- or C282Y-HFE using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. Recovered RNA was quantified by its absorbance at 260nm. cDNA was prepared by reverse transcription of isolated RNA using the RT2 First Strand Kit (SABiosciences, now part of Qiagen). For qRT-PCR reactions, ~4ng cDNA samples were added to a commercially available 96 well Lipoprotein signaling and cholesterol metabolism gene array containing primer pairs specific for the 96 genes involved in cholesterol metabolism (cat # PAHS-080Z) and reactions carried out according to directions provided with the RT2 SYBR® Green/ROX™ qPCR Master Mix Kit (SABiosciences, Frederick, MD). Gene expression of each of the 96 genes was monitored using an ABI 7800 (Applied Biosystems, USA). Data was normalized to expression of mRNA for actin, GAPDH, and ribosomal protein L13A.

182 Cell survival:

WT- and C282Y-HFE cells differ in morphology and proliferation with the latter being larger. In order to treat both cell types when they were approximately equally confluent, 3X104 WT- and 1.5X104 C282Y-HFE cells were seeded per well in a 96 well plate and allowed to grow for 2 days in regular medium (as described above). Cells were then treated with different concentrations of simvastatin (Sigma; S6196) or sphingosine kinase inhibitor (SKI; SKI 178 from Dr. Jong Yun (Hengst et al., 2010; Madhunapantula et al., 2012) or a combination of both. Cells were treated for 24-48 hr in medium containing 1% FBS. Cell survival was measured at 24 and 48 hr time points using the MTS assay according to the manufacturer’s protocol (Promega).

Statistical Analysis:

Data were analyzed using the unpaired Student t-test as appropriate. To compare different treatments, one-way ANOVA followed by the Newman Keuls test was used. Data are shown as percentage of the untreated control. Each experiment was repeated at least three times with 3-8 replicates.

Results

Effect of C282Y-HFE on total cholesterol content

To study the effects of expression of C282Y-HFE on cells, SH-SY5Y cells transfected to stably express WT- or C282Y-HFE protein (Lee et al., 2007b) were used. We previously found that C282Y-HFE expressing cells had significantly increased cell growth and altered morphology compared to WT-HFE cells (Ali-Rahmani et al., 2011). We also showed that while WT-HFE was found in lipid microdomains/rafts, regions of the plasma membrane enriched in cholesterol, glycosphingolipids and signal transduction molecules, C282Y-HFE, was not (Ali-Rahmani et al., 2011). In fact, others have shown that C282Y- HFE stays in the golgi complex and does not localize in the plasma membrane (Waheed et al., 1997). Previous studies indicated that expression of C282Y-HFE was accompanied by changes in glycosphingolipid expression (Ali-Rahmani et al., 2011), which led us to

183 hypothesize that it might also affect metabolism of cholesterol, another lipid enriched in rafts. Biochemical assessment of cholesterol content using a cholesterol assay kit indicated that C282Y-HFE expressing cells had ~18% more cholesterol than WT-HFE cells (Figure 4.1A).

Effect of C282Y-HFE on expression of genes affecting cholesterol metabolism

To investigate the underlying cause(s) for the increased cholesterol content in C282Y- HFE expressing cells relative to those expressing WT-HFE, the expression of 96 key genes encoding proteins involved in cholesterol metabolism was ascertained using a Lipoprotein signaling pathway focused gene array. Expression of a number of genes was altered in C282Y-HFE cells relative to WT-HFE cells (Table1). Only genes with a 4-fold or greater change in expression were considered to be significantly affected. NPC1L1 and CXCL16, the two most highly over-expressed genes (>300 fold), encode the proteins Niemann-Pick disease, type C1 gene-like 1 (functions in uptake of cholesterol by cells) and chemokine ligand 16 (induces cell migration), respectively. Genes with greater than 100 fold upregulation were those encoding apolipoprotein D (APOD), collectin subfamily member 12 (COLEC12), and low density lipoprotein-related protein 1B (LRP1B). While a number of genes were upregulated, expression of cytochrome P450, family 7, subfamily B, polypeptide 1(CYP7B1) was decreased 98-fold in C282Y-HFE cells. While these changes could affect cholesterol metabolism, the genes affected were primarily involved in transport and uptake. For example overexpression of NPC1L1 would be expected to enhance uptake of cholesterol by C282Y-HFE-expressing cells and may have contributed to our finding of increased cholesterol in them.

Effect of cholesterol inhibition on cell survival

Previous studies have shown that C282Y-HFE cells are resistant to chemotherapeutic drugs such as doxorubicin and agents causing differentiation such as retinoic acid (Lee et al., 2010). To determine whether the elevation in cholesterol in C282Y-HFE cells was

184 responsible for their enhanced proliferation and resistance to therapy, we measured the effect of simvastatin (inhibitor of HMGCoAR activity) on cell survival. Both WT- and C282Y-HFE expressing cells were treated with varying concentrations of simvastatin and cell survival assessed 24 and 48 hr later. At 24 hr, no significant difference was observed in survival of cells exposed to either 10 or 20µM simvastatin. However, a greater reduction in survival was observed for WT-HFE cells treated with 40µM simvastatin (~75%) and C282Y-HFE cells (~50%) relative to the untreated WT and C282Y-HFE controls, respectively (Figure 4.2A). After 48 hr, 10µM simvastatin was found to reduce survival of WT-HFE cells by ~20% while 20µM simvastatin was needed to have the same effect on C282Y-HFE cell survival (Figure 4.2B), indicating a genotype-dependent response to treatment with simvastatin on cell survival. These observations support the hypothesis that C282Y-HFE expressing cells may be more efficient at obtaining cholesterol from serum in the growth medium than WT-HFE cells.

Effect of inhibition of sphingosine kinase on cell survival

Previously we found a significant elevation in mRNA, protein expression, and activity of sphingosine kinase 1 (SphK1) in C282Y-HFE cells relative to WT-HFE cells (Ali- Rahmani et al., 2011; Lavieu et al., 2006; Snider et al., 2010; Spiegel and Milstien, 2002). A number of studies have shown association of increased cell proliferation and cell migration with elevation in SphK1 (Bao et al., 2011; Nagahashi et al., 2012; Pan et al., 2011). Therefore, we hypothesized that inhibition of SphK1 would decrease survival of C282Y-HFE cells. Cells were treated with different concentrations of sphingosine kinase inhibitor (SKI) for 24 and 48 hr and survival assessed using the MTS assay. While a significant reduction in survival of WT- and C282Y-HFE cells was observed after 24 and 48 hr of treatment relative to untreated controls (Figure 4.3A & B), no significant differences were observed between WT- and C282Y-HFE cells treated the same way for 24 hr. After 48 hr of treatment, survival of WT-HFE cells was 25% and that of C282Y- HFE cells 55-65% (Figure 4.3A). Interestingly, no significant change in survival of C282Y-HFE cells was observed between 24 and 48 hr of exposure to the sphingosine kinase inhibitor (p>0.05) (Figure 4.3A & B).

185

Effect of exposure to both SKI and simvastatin on cell survival

Given that treatment of C282Y-HFE cells with either simvastatin or SKI alone showed a resistant phenotype, effect of combination of both drugs on the survival of C282Y-HFE cells was tested. Survival of cells treated with both SKI and simvastatin (whether 10µM or 40µM) was significantly (p<0.001) reduced for WT-HFE cells after 24 hr relative to that seen when they were treated with SKI alone and by 48 hr that reduction was markedly enhanced (Fig 4.4A & B and 4.5A & B). The response of C282Y-HFE cells to the combination treatment did not change significantly after treatment with SKI and 10µM of simvastatin for 24 hr or 48 hr (Fig 4.4 C & D). However, addition of 40µM simvastatin along with different concentrations of SKI showed a significant (p<0.01) reduction in survival of C282Y-HFE cells than seen in cells treated with SKI alone after 24 or 48 hr (Fig 4.5 C & D). These data show that WT-HFE cells are more sensitive to SKI, simvastatin alone or combined than C282Y-HFE cells.

Discussion

An extensive literature search revealed only two studies that looked at the effect of expression of C282Y-HFE on cholesterol. Those studies indicated that carriers of C282Y-HFE had low serum total and LDL cholesterol (Adams et al., 2009; Pankow et al., 2008). Our findings that human neuroblastoma cells transfected to express C282Y- HFE had significant changes in expression of a number of genes in the cholesterol pathway are consistent with the human studies. Because a number of the significantly upregulated genes are involved in cholesterol uptake and transport, it can be predicted that the increased level of cholesterol in C282Y-HFE expressing cells could account for clearance of cholesterol from the circulation and uptake into cells of individuals expressing the C282Y-HFE variant.

A prime example of a gene involved in cholesterol transport whose expression was significantly up-regulated in C282Y-HFE cells is NPC1L1 (increased 333-fold). It

186 encodes the Niemann-Pick C1-Like 1 (NPC1L1) transmembrane protein that functions in uptake of cholesterol via clathrin-mediated endocytosis and is regulated by cellular cholesterol content (Brown et al., 2007; Ge et al., 2008; Petersen et al., 2008; Yu et al., 2006).

Additional up-regulated genes involved in cholesterol transport and their functions include the following: CXCL16 (increased 306-fold) which encodes CXCL16/SR-PSOX, an interferon-γ–regulated chemokine found in both trans-membrane and soluble forms, and can function as a scavenger for oxidized low-density lipoproteins (Matloubian et al., 2000; Shimaoka et al., 2000; Wuttge et al., 2004). LRP1B (increased 198-fold) encodes the low density lipoprotein receptor related protein 1B (LRP1B) which interacts with the cytoplasmic tail of the LDL receptor and mediates cholesterol uptake (Li et al., 2002). COLEC12 gene (increased 177-fold) encodes a member of the C-lectin family of proteins. This protein is known to function as a scavenger receptor and can recognize oxidized phospholipids and lipoprotein . APOD encodes for apolipoprotein D (APOD, increased 161-fold) which is associated with high density lipoproteins and interacts with lecithin cholesterol acyltransferase (Holmquist, 1989). The APOE gene encodes apolipoprotein E (APOE, increased 71-fold) which interacts with members of low-density lipoprotein receptor family and plays a key role in lipid transport in the plasma and in the central nervous system (Elshourbagy et al., 1985; Hatters et al., 2006; Ladu et al., 2000). Collectively, these changes in gene expression seen in the C282Y- HFE cells indicate disruption of cholesterol homeostasis, particularly, an increase in cholesterol uptake which may be the reason for the elevated cholesterol content found in these cells.

C282Y-HFE, cholesterol, and cancer risk

As indicated previously, individuals that express C282Y-HFE have an increased risk of breast, colorectal, and prostate cancer relative to WT-HFE carriers. It is possible that this increased risk is due to the C282Y-HFE variant-induced alterations in lipid metabolism. In previous work we found that C282Y-HFE cells expressed significantly more S1P (cell

187 growth promoter) and significantly fewer gangliosides (lipid raft components) than WT- HFE cells (Ali-Rahmani et al., 2011). Here we have shown an elevation in the cholesterol content of C282Y-HFE cells. Cholesterol is an essential membrane component that plays an important role in maintaining membrane integrity and fluidity (Silvius, 2003). It is also critical for formation of lipid rafts where it serves as a spacer between the hydrocarbon chains of sphingolipids (London and Brown, 2000; Simons and Toomre, 2000). Therefore, it has been hypothesized that alterations in the cholesterol content of cells changes the composition and properties of lipid rafts. In fact, studies have shown that depletion of cholesterol from the plasma membrane causes disruption of lipid rafts and movement of raft-associated proteins into non-raft areas of the membrane. The change in membrane environment can affect protein function and in turn, cell survival (Scheel-Toellner et al., 2002; Simons and Toomre, 2000).

Cholesterol inhibition-potential therapy for cancer

Considering the role of cholesterol in cell growth and cancer promotion, it is logical to predict that cholesterol inhibition/depletion could be beneficial for patients with a cancer that accumulates cholesterol such as breast and prostate. Support for this hypothesis is provided by studies showing a reduction in the incidence of prostate cancer in patients on long-term statin therapy (Graaf et al., 2004)( (Platz et al., 2006). In addition, depletion of cholesterol from human prostate cancer cells harboring high amounts of cholesterol resulted in increased sensitivity to apoptosis (Li et al., 2006). Cholesterol depletion using effects that were reversed by cholesterol repletion. Similar effects were achieved in cells by treatment with simvastatin (Li et al., 2006) to inhibit HMGCoAR, the rate limiting enzyme in cholesterol synthesis. Use of statins has been reported to be beneficial for the treatment of some cancers as well. Though conflicting data have been found for the effect of statin treatment on the risk of cancer (Dale et al., 2006), several reports have consistently documented their efficacy, especially for prostate cancer (Solomon and Freeman, 2008). Our results indicate that while treatment of cholesterol-enriched C282Y- HFE cells with simvastatin decreased their survival they were less sensitive to

188 simvastatin than WT-HFE cells. The reduced sensitivity of C282Y-HFE cells to simvastatin may reflect their ability to obtain cholesterol from the medium via upregulated expression of NPC1L1. These observations may help to explain the apparently conflicting findings seen in human studies of statin treatment for certain cancers. It is possible that when the same dose was given to all subjects in a clinical study without the knowledge of their HFE genotype, the inclusion of C282Y-HFE carriers in the subject pool affected the outcome. Further prospective studies are warranted to ascertain whether individuals with C282Y-HFE have elevated levels of cholesterol and whether a higher dose of statin treatment might be a better treatment option for them. Nonetheless, our findings along with other studies suggest that cancers whose cells exhibit an upregulation in expression of enzymes involved in cholesterol synthesis may be more responsive to statin therapy than those with elevated cholesterol levels due to overexpression of proteins that enhance its uptake by cells.

Association of cholesterol metabolizing genes with cancer phenotype

Some of the genes we found elevated in C282Y-HFE cells, have been shown to play a role in cancer phenotypes by others. For example, CXCL16 was found to stimulate cell proliferation, chemotaxis, and angiogenesis (Darash-Yahana et al., 2009; Lu et al., 2008; Rabquer et al., 2011). It is found in both trans-membrane and soluble forms and is known to bind the chemokine receptor, CXCR6, and guide migration of T-cells (Wuttge et al., 2004). Elevated expression of CXCL16 has been found in atherosclerotic lesions, suggesting a link between inflammation and lipid metabolism (Darash-Yahana et al., 2009; Shimaoka et al., 2000; Wuttge et al., 2004). In addition, elevated expression of CXCL16 was observed in several prostate cancer cell lines with much higher expression in more aggressive PC3 cells relative to less aggressive LNCaP cells as well as in prostate tumors from humans (Lu et al., 2008). CXCL16 was also found to induce prostate cancer cell migration and invasion in a dose-dependent manner. We found significant upregulation of mRNA and protein expression, as well as, soluble CXCL16 in C282Y-HFE cells. These data support an association between function of CXCL16 and phenotype of C282Y-HFE cells.

189 A role for NPC1L1 in cancer is unknown but considering the role of accumulated cholesterol in cancer promotion and known function of NPC1L1 in cholesterol uptake, future studies focusing on this connection may provide additional targets to inhibit cholesterol-dependent cancer progression.

Though no significant association has been found between isoforms of APOE and cancer, a few studies suggest that possibility. For example, macrophages surrounding the human colonic adenocarcinoma cells have been shown to overexpress APOE (Niemi et al., 2002; Niemi et al., 2000). In addition, prostate cancer patients carrying the APOE3 or APOE4 isoform had high serum cholesterol relative to the ones with benign prostate hyperplasia (Niemi et al., 2000).

ApoD, a component of high-density lipoproteins (HDL), plays a significant role in cholesterol metabolism and its abnormal expression is implicated in different metabolic syndromes. For example, Do Carmo et al. reported induction of ApoD expression in NIH/3T3 cells in response to cellular stress, regardless of whether the stress was caused by lipopolysaccharide (LPS) stimulation, H2O2 treatment or UV-light irradiation (Do Carmo et al., 2007). In addition, elevated ApoD production was detected in liver tumors resected from hepatocellular carcinoma (Vizoso et al., 2007), as well as in invasive carcinoma of the breast (Gonzalez et al., 2007; Soiland et al., 2009). Collectively, these studies support the relevance of observed changes in C282Y-HFE cells to cancer phenotype.

In conclusion, given that cholesterol is accumulated in several tumors and is implicated in induction of cell survival pathways via activation of AKT, findings from our studies suggest that the increase in total cholesterol and S-1-P in C282Y-HFE cells could contribute to AKT activation in these cells. Evidence for the involvement of cholesterol in AKT activation and cell proliferation and tumor growth, combined with our findings of elevation of cholesterol content and cell proliferation in C282Y-HFE cells indicates a potential mechanism to explain the higher incidence and severity of breast, prostate, and colon cancers in carriers of C282Y-HFE which warrants further investigation. Our findings, consistent with others, support a role for the C282Y-HFE

190 variant in malignancies and argue for stratification of subjects by HFE genotype to access disease susceptibility and to develop effective therapeutics.

191 Table 4.1: Effect of C282Y-HFE variant on the expression of genes involved in cholesterol metabolism relative to WT-HFE-expressing cells

Pools of total RNA from WT- or C282Y-HFE stably transfected cells were analyzed using the GEArray human target specific gene array. Changes in mRNA levels are expressed as  (fold increase),  (fold decrease). The average changes in expression (> 4 folds) are shown.

192 Figure 4.1: Effect of expression of C282Y-HFE on cholesterol content relative to WT- HFE expressing cells

A) Cholesterol content of SH-SY5Y cells expressing WT- or C282Y-HFE. Lipids were extracted from equal number of cells (106) of each cell type and cholesterol content - and C282Y-HFE cells.

WT-HFE C282Y-HFE

193 Figure 4.2: Effect of simvastatin treatment on survival of WT- and C282Y-HFE cells. Both WT- and C282Y-HFE expressing cells were treated with 10, 20, or 40 µM simvastatin for either 24 or 48 hours and cell survival was measured by the MTS assay. Untreated cells served as control. A) Cell survival significantly decreased after treatment with 40uM of statin for 24 hrs for both cell types relative to the untreated controls *** P < 0.001, n=8. The comparison of 40uM treated C282Y cells with WT-HFE cells treated analogously is indicated by ϯϯϯ p<0.001 Decreased cell survival was observed for WT- HFE expressing cells relative to expressing C282Y-HFE after treatment with simvastatin for 24 hrs. B) Treatments were done analogously but cells survival was measured after 48 hours. A.

194 Figure 4.3 Effect of sphingosine kinase inhibitor (SKI) on cell survival A) SKI alone 24 hr B) SKI alone 48 hr. Both WT- and C282Y-HFE expressing cells were treated with 1, 5, 10, or 20 µM SKI for either 24 or 48 hours and cell survival was measured by the MTS assay. Untreated cells served as control. A) Relative to the untreated controls, survival of both WT- and C282Y-HFE cells was significantly decreased after treatment with 1, 5, 10, or 20 µM SKI of SKI for 24 hrs and B) 48 hours and indicated by *** P < 0.001, n=8. The comparison of each treatment between WT- and C282Y-HFE cells, where significant, is shown by ϯ p<0.05; ϯϯϯ p<0.001. After 24 hr treatment, with 5 or 10uM of SKI, C282Y-HFE cells are less sensitive to treatment than WT-HFE cells treated the same as shown in A. However, after 48 hrs, C282Y-HFE cells are more resistant ti SKI than WT-HFE cells at all concentration used and shown in panel B. A.

195 Figure 4.4: Effect of combination treatment of simvastatin and sphingosine kinase inhibitor on cell survival treatment- 10uM of statin Cells were treated with either 1,5, 10, or 20 uM of SKI alone, 10uM of statin alone, or conbination of both for 24 or 48 hrs before assaying for survival. A) WT-HFE cells, with single or combination treatment, had decreased survival relative to the untreated controls (*** p<0.001) or relative to 10uM statin treated cells (ʈʈʈ p<0.001). B) WT-HFE were treated as described in A, except cell survival was measure after treating cells for 48 hrs C) Effects of treatments as described above on C282Y-HFE cells after 24 hr and D) 48 hr.*** indicates comparison of treatment group with untreated controls. ʈʈʈ indicates comparison of treatment group with statin treatment alone (last bar in the graph). Wt-0 is untreated control; WT-1 is treated with 1uM SKI; WT-1+10 is treated with 1uM SKI plus 10uM statin. A. B.

C. D.

196 Figure 4.5: Effect of combination treatment of simvastatin and sphingosine kinase inhibitor on cell survival treatment- 40uM of statin Cells were treated with either 1,5, 10, or 20 uM of SKI alone, 40uM of statin alone, or conbination of both for 24 or 48 hrs before assaying for survival. A) WT-HFE cells, with single or combination treatment, had decreased survival relative to the untreated controls (*** p<0.001) or relative to 40uM statin treated cells (ʈʈʈ p<0.001). B) WT-HFE were treated as described in A, except cell survival was measure after treating cells for 48 hrs C) Effects of treatments as described above on C282Y-HFE cells after 24 hr and D) 48 hr.*** indicates comparison of treatment group with untreated controls. ʈʈʈ indicates comparison of treatment group with statin treatment alone (last bar in the graph). Wt-0 is untreated control; WT-1 is treated with 1uM SKI; WT-1+40 is treated with 1uM SKI plus 40uM statin. A.

197 C.

198 References

Adam, R.M., N.K. Mukhopadhyay, J. Kim, D. Di Vizio, B. Cinar, K. Boucher, K.R. Solomon, and M.R. Freeman. 2007. Cholesterol sensitivity of endogenous and myristoylated Akt. Cancer Research. 67:6238-6246.

Adams, P.C., J.S. Pankow, J.C. Barton, R.T. Acton, C. Leiendecker-Foster, G.D. McLaren, M. Speechley, and J.H. Eckfeldt. 2009. HFE C282Y homozygosity is associated with lower total and low-density lipoprotein cholesterol: The hemochromatosis and iron overload screening study. Circ Cardiovasc Genet. 2:34-7.

Ali-Rahmani, F., J.A. Hengst, J.R. Connor, and C.L. Schengrund. 2011. Effect of HFE Variants on Sphingolipid Expression by SH-SY5Y Human Neuroblastoma Cells. Neurochem Res.

Ba, Q., M. Hao, H. Huang, J. Hou, S. Ge, Z. Zhang, J. Yin, R. Chu, H. Jiang, F. Wang, K. Chen, H. Liu, and H. Wang. 2011. Iron deprivation suppresses hepatocellular carcinoma growth in experimental studies. Clin Cancer Res. 17:7625-33.

Bao, M., Z. Chen, Y. Xu, Y. Zhao, R. Zha, S. Huang, L. Liu, T. Chen, J. Li, H. Tu, and X. He. 2011. Sphingosine kinase 1 promotes tumour cell migration and invasion via the S1P/EDG1 axis in hepatocellular carcinoma. Liver Int. 32:331-8.

Bassett, M.L., J.W. Halliday, and L.W. Powell. 1986. Value of Hepatic Iron Measurements in Early Hemochromatosis and Determination of the Critical Iron Level Associated with Fibrosis. Hepatology. 6:24-29.

Bennis, F., G. Favre, F. Legaillard, and G. Soula. 1993. Importance of Mevalonate- Derived Products in the Control of Hmg-Coa Reductase-Activity and Growth of Human Lung Adenocarcinoma Cell-Line A549. International Journal of Cancer. 55:640-645.

199 Brown, J.M., L.L. Rudel, and L.Q. Yu. 2007. NPC1L1 (Niemann-Pick Cl-like 1) mediates sterol-specific unidirectional transport of non-esterified cholesterol in McArdle-RH7777 hepatoma cells. Biochemical Journal. 406:273-283.

Camaschella, C., A. Roetto, and M. De Gobbi. 2002. Genetic haemochromatosis: genes and mutations associated with iron loading. Best Practice & Research Clinical Haematology. 15:261-276.

Cardone, M.H., N. Roy, H.R. Stennicke, G.S. Salvesen, T.F. Franke, E. Stanbridge, S. Frisch, and J.C. Reed. 1998. Regulation of cell death protease caspase-9 by phosphorylation. Science. 282:1318-1321.

Dale, K.M., C.I. Coleman, N.N. Henyan, J. Kluger, and C.M. White. 2006. Statins and cancer risk: a meta-analysis. JAMA. 295:74-80.

Darash-Yahana, M., J.W. Gillespie, S.M. Hewitt, Y.Y.K. Chen, S. Maeda, I. Stein, S.P. Singh, R.B. Bedolla, A. Peled, D.A. Troyer, E. Pikarsky, M. Karin, and J.M. Farber. 2009. The Chemokine CXCL16 and Its Receptor, CXCR6, as Markers and Promoters of Inflammation-Associated Cancers. PLoS One. 4.

Datta, S.R., H. Dudek, X. Tao, S. Masters, H.A. Fu, Y. Gotoh, and M.E. Greenberg. 1997. Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell. 91:231-241.

Do Carmo, S., L.C. Levros, Jr., and E. Rassart. 2007. Modulation of apolipoprotein D expression and translocation under specific stress conditions. Biochim Biophys Acta. 1773:954-69.

Dongiovanni, P., A.L. Fracanzani, G. Cairo, C.P. Megazzini, S. Gatti, R. Rametta, S. Fargion, and L. Valenti. 2010. Iron-Dependent Regulation of MDM2 Influences p53 Activity and Hepatic Carcinogenesis. American Journal of Pathology. 176:1006-1017.

200 Duncan, R.E., A. El-Sohemy, and M.C. Archer. 2004. Mevalonate promotes the growth of tumors derived from human cancer cells in vivo and stimulates proliferation in vitro with enhanced cyclin-dependent kinase-2 activity. Journal of Biological Chemistry. 279:33079-33084.

El-Sohemy, A., and M.C. Archer. 2000. Inhibition of N-methyl-N-nitrosourea- and 7,12- dimethylbenz[alpha]anthracene-induced rat mammary tumorigenesis by dietary cholesterol is independent of Ha-ras mutations. Carcinogenesis. 21:827-831.

Elhyany, S., E. Assa-Kunik, S. Tsory, T. Muller, S. Fedida, S. Segal, and D. Fishman. 2004. The integrity of cholesterol-enriched microdomains is essential for the constitutive high activity of protein kinase B in tumour cells. Biochemical Society Transactions. 32:837-839.

Elshourbagy, N.A., W.S. Liao, R.W. Mahley, and J.M. Taylor. 1985. Apolipoprotein E mRNA is abundant in the brain and adrenals, as well as in the liver, and is present in other peripheral tissues of rats and marmosets. Proc Natl Acad Sci U S A. 82:203-7.

Fargion, S., L. Valenti, and A.L. Fracanzani. 2010. Hemochromatosis gene (HFE) mutations and cancer risk: expanding the clinical manifestations of hereditary iron overload. Hepatology. 51:1119-21.

Freeman, M.R., and K.R. Solomon. 2004. Cholesterol and prostate cancer. Journal of Cellular Biochemistry. 91:54-69.

201 Gannon, P.O., S. Medelci, C. Le Page, M. Beaulieu, D.M. Provencher, A.M. Mes- Masson, and M.M. Santos. 2011. Impact of hemochromatosis gene (HFE) mutations on epithelial ovarian cancer risk and prognosis. International Journal of Cancer. 128:2326-2334.

Ge, L., J. Wang, W. Qi, H.H. Miao, J. Cao, Y.X. Qu, B.L. Li, and B.L. Song. 2008. The cholesterol absorption inhibitor ezetimibe acts by blocking the sterol-induced internalization of NPC1L1. Cell Metabolism. 7:508-519.

Gonzalez, L.O., M.D. Corte, S. Junquera, M. Bongera, J.C. Rodriguez, and F.J. Vizoso. 2007. Expression of androgen receptor and two androgen-induced proteins (apolipoprotein D and pepsinogen C) in ductal carcinoma in situ of the breast. Histopathology. 50:866-74.

Graaf, M.R., A.B. Beiderbeck, A.C.G. Egberts, D.J. Richel, and H.J. Guchelaar. 2004. The risk of cancer in users of statins. Journal of Clinical Oncology. 22:2388- 2394.

Graf, E., and J.W. Eaton. 1985. Dietary Suppression of Colonic-Cancer - Fiber or Phytate. Cancer. 56:717-718.

Greenberg, G. 1976. Sarcoma after intramuscular iron injection. BMJ. 1:1508-1509.

Hatters, D.M., C.A. Peters-Libeu, and K.H. Weisgraber. 2006. Apolipoprotein E structure: insights into function. Trends in Biochemical Sciences. 31:445-454.

Hellerbrand, C., A. Poppl, A. Hartmann, J. Scholmerich, and G. Lock. 2003. HFE C282Y Heterozygosity in Hepatocellular Carcinoma: Evidence for an Increased Prevalence. Clinical Gastroenterology and Hepatology. 1:279-284.

Hengst, J.A., X.J. Wang, U.H. Sk, A.K. Sharma, S. Amin, and J.K. Yun. 2010. Development of a sphingosine kinase 1 specific small-molecule inhibitor. Bioorganic & Medicinal Chemistry Letters. 20:7498-7502.

202 Hill, M.M., J.H. Feng, and B.A. Hemmings. 2002. Identification of a plasma membrane raft-associated PKB Ser473 kinase activity that is distinct from ILK and PDK1. Current Biology. 12:1251-1255.

Holmquist, L. 1989. Separation of Free and Apolipoprotein D-Associated Human-Plasma Lecithin - Cholesterol Acyltransferase. Journal of Biochemical and Biophysical Methods. 19:93-104.

Kane, L.P., V.S. Shapiro, D. Stokoe, and A. Weiss. 1999. Induction of NF-kappa B by the Akt PKB kinase. Current Biology. 9:601-604.

Kazantzis, G. 1981. Role of Cobalt, Iron, Lead, Manganese, Mercury, Platinum, Selenium, and Titanium in Carcinogenesis. Environmental Health Perspectives. 40:143-161.

Kolanjiappan, K., C.R. Ramachandran, and S. Manoharan. 2003. Biochemical changes in tumor tissues of oral cancer patients. Clinical Biochemistry. 36:61-65.

Ladu, M.J., C. Reardon, L. Van Eldik, A.M. Fagan, G. Bu, D. Holtzman, and G.S. Getz. 2000. Lipoproteins in the central nervous system. Ann N Y Acad Sci. 903:167-75.

Lauret, E., M. Rodriguez, S. Gonzalez, A. Linares, A. Lopez-Vazquez, J. Martinez-Borra, L. Rodrigo, and C. Lopez-Larrea. 2002. HFE gene mutations in alcoholic and virus-related cirrhotic patients with hepatocellular carcinoma. American Journal of Gastroenterology. 97:1016-1021.

Lavieu, G., F. Scarlatti, G. Sala, S. Carpentier, T. Levade, R. Ghidoni, J. Botti, and P. Codogno. 2006. Regulation of autophagy by sphingosine kinase 1 and its role in cell survival during nutrient starvation. Journal of Biological Chemistry. 281:8518-8527.

Lee, S.Y., S. Liu, R.M. Mitchell, B. Slagle-Webb, Y.S. Hong, J.M. Sheehan, and J.R. Connor. 2010. HFE polymorphisms influence the response to chemotherapeutic agents via induction of p16INK4A. Int J Cancer.

203 Lee, S.Y., S.M. Patton, R.J. Henderson, and J.R. Connor. 2007a. Consequences of expressing mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking endogenous HFE. FASEB J. 21:564-76.

Lee, S.Y., S.M. Patton, R.J. Henderson, and J.R. Connor. 2007b. Consequences of expressing mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking endogenous HFE. FASEB J. 21:564-576.

Li, H.Y., F.R. Appelbaum, C.L. Willman, R.A. Zager, and D.E. Banker. 2003. Cholesterol-modulating agents kill acute myeloid leukemia cells and sensitize them to therapeutics by blocking adaptive cholesterol responses. Blood. 101:3628-3634.

Li, Y.C., M.J. Park, S.K. Ye, C.W. Kim, and Y.N. Kim. 2006. Elevated levels of cholesterol-rich lipid rafts in cancer cells are correlated with apoptosis sensitivity induced by cholesterol-depleting agents. American Journal of Pathology. 168:1107-1118.

Li, Y.H., J.M. Knisely, W.Y. Lu, L.M. McCormick, J.Y. Wang, J. Henkin, A.L. Schwartz, and G.J. Bu. 2002. Low density lipoprotein (LDL) receptor-related protein 1B impairs urokinase receptor regeneration on the cell surface and inhibits cell migration. Journal of Biological Chemistry. 277:42366-42371.

London, E., and D.A. Brown. 2000. Insolubility of lipids in Triton X-100: physical origin and relationship to sphingolipid/cholesterol membrane domains (rafts). Biochimica Et Biophysica Acta-Biomembranes. 1508:182-195.

Lu, Y., J.H. Wang, Y. Xu, A.E. Koch, Z. Cai, X. Chen, D.L. Galson, R.S. Taichman, and J. Zhang. 2008. CXCL16 functions as a novel chemotactic factor for prostate cancer cells in vitro. Molecular Cancer Research. 6:546-554.

204 Lund, E.K., S.G. Wharf, S.J. Fairweather-Tait, and I. Johnson. 1999. Oral ferrous sulfate supplements increase the free radical-generating capacity of feces from healthy volunteers. American Journal of Clinical Nutrition. 69:250-255.

Lund, E.K., S.G. Wharf, S.J. Fairweather-Tait, and I.T. Johnson. 1998. Increases in the concentrations of available iron in response to dietary iron supplementation are associated with changes in crypt cell proliferation in rat large intestine. Journal of Nutrition. 128:175-179.

Madhunapantula, S.V., J. Hengst, R. Gowda, T.E. Fox, J.K. Yun, and G.P. Robertson. 2012. Targeting sphingosine kinase-1 to inhibit melanoma. Pigment Cell & Melanoma Research. 25.

Matloubian, M., A. David, S. Engel, J.E. Ryan, and J.G. Cyster. 2000. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nature Immunology. 1:298-304.

Milman, N., P. Pedersen, T.A. Steig, K.E. Byg, N. Graudal, and K. Fenger. 2001. Clinically overt hereditary hemochromatosis in Denmark 1948-1985: epidemiology, factors of significance for long-term survival, and causes of death in 179 patients. Annals of Hematology. 80:737-744.

Miyata, Y., S. Koga, M. Nishikido, T. Hayashi, and H. Kanetake. 2001. Relationship between serum ferritin levels and tumour status in patients with renal cell carcinoma. Bju International. 88:974-977.

Mizukami, S., R. Ichimura, S. Kemmochi, L.Y. Wang, E. Taniai, K. Mitsumori, and M. Shibutani. 2010. Tumor promotion by copper-overloading and its enhancement by excess iron accumulation involving oxidative stress responses in the early stage of a rat two-stage hepatocarcinogenesis model. Chemico-Biological Interactions. 185:189-201.

205 Nagahashi, M., S. Ramachandran, E.Y. Kim, J.C. Allegood, O.M. Rashid, A. Yamada, R. Zhao, S. Milstien, H. Zhou, S. Spiegel, and K. Takabe. 2012. Sphingosine-1- phosphate produced by sphingosine kinase 1 promotes breast cancer progression by stimulating angiogenesis and lymphangiogenesis. Cancer Res. 72:726-35.

Nelson, R.L. 2001. Iron and colorectal cancer risk: Human studies. Nutrition Reviews. 59:140-148.

Niederau, C., R. Fischer, A. Purschel, W. Stremmel, D. Haussinger, and G. Strohmeyer. 1996. Long-term survival in patients with hereditary hemochromatosis. Gastroenterology. 110:1107-1119.

Niemi, M., T. Hakkinen, T.J. Karttunen, S. Eskelinen, K. Kervinen, M.J. Savolainen, J. Lehtola, J. Makela, S. Yla-Herttuala, and Y.A. Kesaniemi. 2002. Apolipoprotein E and colon cancer. Expression in normal and malignant human intestine and effect on cultured human colonic adenocarcinoma cells. Eur J Intern Med. 13:37- 43.

Niemi, M., K. Kervinen, H. Kiviniemi, O. Lukkarinen, A.P. Kyllonen, M. Apaja- Sarkkinen, M.J. Savolainen, M.I. Kairaluoma, and Y.A. Kesaniemi. 2000. Apolipoprotein E phenotype, cholesterol and breast and prostate cancer. J Epidemiol Community Health. 54:938-9.

Oh, H.Y., E.J. Lee, S. Yoon, B.H. Chung, K.S. Cho, and S.J. Hong. 2007. Cholesterol level of lipid raft microdomains regulates apoptotic cell death in prostate cancer cells through EGFR-mediated Akt and ERK signal transduction. Prostate. 67:1061-1069.

Osawa, Y., H. Uchinami, J. Bielawski, R.F. Schwabe, Y.A. Hannun, and D.A. Brenner. 2005. Roles for C16-ceramide and sphingosine 1-phosphate in regulating hepatocyte apoptosis in response to tumor necrosis factor-alpha. J Biol Chem. 280:27879-87.

206 Osborne, N.J., L.C. Gurrin, K.J. Allen, C.C. Constantine, M.B. Delatycki, C.E. McLaren, D.M. Gertig, G.J. Anderson, M.C. Southey, J.K. Olynyk, L.W. Powell, J.L. Hopper, G.G. Giles, and D.R. English. 2010. HFE C282Y Homozygotes Are at Increased Risk of Breast and Colorectal Cancer. Hepatology. 51:1311-1318.

Pan, J., Y.F. Tao, Z. Zhou, B.R. Cao, S.Y. Wu, Y.L. Zhang, S.Y. Hu, W.L. Zhao, J. Wang, G.L. Lou, Z. Li, X. Feng, and J. Ni. 2011. An novel role of sphingosine kinase-1 (SPHK1) in the invasion and metastasis of esophageal carcinoma. J Transl Med. 9:157.

Panka, D.J., T. Mano, T. Suhara, K. Walsh, and J.W. Mier. 2001. Phosphatidylinositol 3- kinase/Akt activity regulates c-FLIP expression in tumor cells. Journal of Biological Chemistry. 276:6893-6896.

Partovian, C., and M. Simons. 2004. Regulation of protein kinase B/Akt activity and Ser(473) phosphorylation by protein kinase C alpha in endothelial cells. Cellular Signalling. 16:951-957.

Petersen, N.H., N.J. Faegeman, L.Q. Yu, and D. Wustner. 2008. Kinetic imaging of NPC1L1 and sterol trafficking between plasma membrane and recycling endosomes in hepatoma cells. Journal of Lipid Research. 49:2023-2037.

Platz, E.A., M.F. Leitzmann, K. Fisvanathan, E.B. Rimm, M.J. Stampfer, W.C. Willett, and E. Giovannucci. 2006. Statin drugs and risk of advanced prostate cancer. Journal of the National Cancer Institute. 98:1819-1825.

207 Porstmann, T., B. Griffiths, Y.L. Chung, O. Delpuech, G. Jr, J. Downward, and A. Schulze. 2005. PKB/Akt induces transcription of enzymes involved in cholesterol and fatty acid biosynthesis via activation of SREBP. Oncogene. 24:6465-6481.

Rabquer, B.J., P.S. Tsou, Y. Hou, E. Thirunavukkarasu, G.K. Haines, A.J. Impens, K. Phillips, B. Kahaleh, J.R. Seibold, and A.E. Koch. 2011. Dysregulated expression of MIG/CXCL9, IP-10/CXCL10 and CXCL16 and their receptors in systemic sclerosis. Arthritis Research & Therapy. 13.

Richmond, H.G. 1959. Induction of Sarcoma in the Rat by Iron-Dextran Complex. British Medical Journal. 1:947-949.

Romashkova, J.A., and S.S. Makarov. 1999. NF-kappa B is a target of AKT in anti- apoptotic PDGF signalling. Nature. 401:86-90.

Sato, Y., Y. Honda, T. Asoh, K. Oizumi, Y. Ohshima, and E. Honda. 1998. Cerebrospinal fluid ferritin in glioblastoma: Evidence for tumor synthesis. Journal of Neuro- Oncology. 40:47-50.

Scheel-Toellner, D., K. Wang, R. Singh, S. Majeed, K. Raza, S.J. Curnow, M. Salmon, and J.M. Lord. 2002. The death-inducing signalling complex is recruited to lipid rafts in Fas-induced apoptosis. Biochemical and Biophysical Research Communications. 297:876-879.

Selby, J.V., and G.D. Friedman. 1988. Epidemiologic Evidence of an Association between Body Iron Stores and Risk of Cancer. International Journal of Cancer. 41:677-682.

Sesink, A.L.A., D.S.M.L. Termont, J.H. Kleibeuker, and R. Van der Meer. 1999. Red meat and colon cancer: The cytotoxic and hyperproliferative effects of dietary heme. Cancer Research. 59:5704-5709.

Shaheen, N.J., L.M. Silverman, T. Keku, L.B. Lawrence, E.M. Rohlfs, C.F. Martin, J. Galanko, and R.S. Sandler. 2003. Association between hemochromatosis (HFE)

208 gene mutation carrier status and the risk of colon cancer. Journal of the National Cancer Institute. 95:154-159.

Shimamura, H., Y. Terada, T. Okado, H. Tanaka, S. Inoshita, and S. Sasaki. 2003. The PI3-kinase-Akt pathway promotes mesangial cell survival and inhibits apoptosis in vitro via NF-kappa B and Bad. Journal of the American Society of Nephrology. 14:1427-1434.

Shimaoka, T., N. Kume, M. Minami, K. Hayashida, H. Kataoka, T. Kita, and S. Yonehara. 2000. Molecular cloning of a novel scavenger receptor for oxidized low density lipoprotein, SR-PSOX, on macrophages. Journal of Biological Chemistry. 275:40663-40666.

Silvius, J.R. 2003. Role of cholesterol in lipid raft formation: lessons from lipid model systems. Biochimica Et Biophysica Acta-Biomembranes. 1610:174-183.

Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nature Reviews Molecular Cell Biology. 1:31-39.

Snider, A.J., K.A.O. Gandy, and L.M. Obeid. 2010. Sphingosine kinase: Role in regulation of bioactive sphingolipid mediators in inflammation. Biochimie. 92:707-715.

Soiland, H., E.A. Janssen, H. Korner, J.E. Varhaug, I. Skaland, E. Gudlaugsson, J.P. Baak, and J.A. Soreide. 2009. Apolipoprotein D predicts adverse outcome in women >or=70 years with operable breast cancer. Breast Cancer Res Treat. 113:519-28.

Solomon, K.R., and M.R. Freeman. 2008. Do the cholesterol-lowering properties of statins affect cancer risk? Trends Endocrinol Metab. 19:113-21.

Spiegel, S., and S. Milstien. 2002. Sphingosine 1-phosphate, a key cell signaling molecule. Journal of Biological Chemistry. 277:25851-25854.

209 Stevens, R.G., R.P. Beasley, and B.S. Blumberg. 1986. Iron-Binding Proteins and Risk of Cancer in Taiwan. Journal of the National Cancer Institute. 76:605-610.

Stevens, R.G., D.Y. Jones, M.S. Micozzi, and P.R. Taylor. 1988. Body Iron Stores and the Risk of Cancer. New England Journal of Medicine. 319:1047-1052.

Stevens, R.G., J.E. Morris, G.A. Cordis, L.E. Anderson, D.W. Rosenberg, and L.B. Sasser. 2003. Oxidative damage in colon and mammary tissue of the HFE- knockout mouse. Free Radical Biology and Medicine. 34:1212-1216.

Stojiljkovic, I., and D. Perkins-Balding. 2002. Processing of heme and heme-containing proteins by bacteria. DNA and Cell Biology. 21:281-295.

Vizoso, F.J., M. Rodriguez, A. Altadill, M.L. Gonzalez-Dieguez, A. Linares, L.O. Gonzalez, S. Junquera, F. Fresno-Forcelledo, M.D. Corte, and L. Rodrigo. 2007. Liver expression of steroid hormones and Apolipoprotein D receptors in hepatocellular carcinoma. World J Gastroenterol. 13:3221-7.

Waheed, A., S. Parkkila, X.Y. Zhou, S. Tomatsu, Z. Tsuchihashi, J.N. Feder, R.C. Schatzman, R.S. Britton, B.R. Bacon, and W.S. Sly. 1997. Hereditary hemochromatosis: effects of C282Y and H63D mutations on association with beta2-microglobulin, intracellular processing, and cell surface expression of the HFE protein in COS-7 cells. Proc Natl Acad Sci U S A. 94:12384-9.

Wu, T.J., C.T. Sempos, J.L. Freudenheim, P. Muti, and E. Smith. 2004. Serum iron, copper and zinc concentrations and risk of cancer mortality in US adults. Annals of Epidemiology. 14:195-201.

210 Wurzelmann, J.I., A. Silver, D.M. Schreinemachers, R.S. Sandler, and R.B. Everson. 1996. Iron intake and the risk of colorectal cancer. Cancer Epidemiology Biomarkers & Prevention. 5:503-507.

Wuttge, D.M., X. Zhou, Y. Sheikine, D. Wagsater, V. Stemme, U. Hedin, S. Stemme, G.K. Hansson, and A. Sirsjo. 2004. CXCL16/SR-PSOX is an interferon-gamma- regulated chemokine and scavenger receptor expressed in atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 24:750-5.

Yu, L.Q., S. Bharadwaj, J.M. Brown, Y.Y. Ma, W. Du, M.A. Davis, P. Michaely, P.S. Liu, M.C. Willingham, and L.L. Rudel. 2006. Cholesterol-regulated translocation of NPC1L1 to the cell surface facilitates free cholesterol uptake. Journal of Biological Chemistry. 281:6616-6624.

Zacharski, L.R., B.K. Chow, P.S. Howes, G. Shamayeva, J.A. Baron, R.L. Dalman, D.J. Malenka, C.K. Ozaki, and P.W. Lavori. 2008. Decreased cancer risk after iron reduction in patients with peripheral arterial disease: Results from a randomized trial. Journal of the National Cancer Institute. 100:996-1002.

Zhuang, L., J. Lin, M. Lu, K.R. Solomon, and M.R. Freeman. 2001. Caveolae-related cholesterol rich lipid-rafts regulate EGF-induced Akt signaling in the LNCaP human prostate cancer cell model. Clinical Cancer Research. 7:3774s-3774s.

Zhuang, L.Y., J. Kim, F. You, R.M. Adam, K.R. Solomon, and M.R. Freeman. 2003. Cholesterol regulates cell survival in human prostate cancer xenografts. Clinical Cancer Research. 9:6178s-6179s.

Zhuang, L.Y., J.Q. Lin, M.L. Lu, K.R. Solomon, and M.R. Freeman. 2002. Cholesterol- rich lipid rafts mediate Akt-regulated survival in prostate cancer cells. Cancer Research. 62:2227-2231.

211 Chapter 5 Summary and Future Perspectives

Introduction:

The main objective of this thesis was to determine whether a relationship exists between iron and cholesterol metabolism that could be of relevance for understanding the underlying mechanism of diseases such as AD and cancer. This goal was achieved by utilizing novel cell culture and mouse models expressing HFE or its variants. The characteristic feature of intracellular and brain iron accumulation resulting from the expression of H63D-HFE allowed us to simultaneously evaluate effects of expression of H63D-HFE and its associated iron overload on lipid metabolism and memory function both of which are known to be disrupted in AD. Results presented in this thesis are the first to establish an association of H63D-HFE with disruption of cholesterol metabolism, neurodegeneration, and memory impairment. Considering the high prevalence (up to 24%) of expression of H63D-HFE in the general population, especially Caucasians, our findings could have a significant impact not only on the understanding of mechanisms underlying memory impairment in carriers of this variant, but results of these findings could also be extrapolated to understand how changes in environmental factors such as diet could influence the risk of dementia.

In this thesis, I have provided evidence that the expression of H63D-HFE in cells and H67D-HFE in mice, was accompanied by a significant reduction in both cell and brain cholesterol. I also found that expression of H67D-HFE was associated with a reduction in expression of synaptic proteins and proteins needed for myelination, an increase in neurodegeneration, and impaired learning and memory. This is the first study to show that in addition to affecting iron homeostasis, a variant of HFE affects cholesterol metabolism and specifically links the H67D-HFE phenotype to decreases in learning and memory and therefore may have implications for AD pathogenesis. Collectively, my data are summarized in figure 5.1.

212 Disruptions in the maintenance of iron and cholesterol metabolism have also been implicated in several cancers. Therefore, in the second part of this work, studies of the effects of expression of C282Y-HFE on lipid (cholesterol and sphingolipid) metabolism and their possible relationship to cancer phenotype are described. Expression of C282Y- HFE altered transcription of a number of enzymes involved in cholesterol and sphingolipid metabolism. Of particular interest were the findings that expression of C282Y-HFE was associated with increased cell proliferation, elevated cholesterol content and elevated expression of the enzyme sphingosine kinase 1(SphK) and its pro-survival metabolite, sphingosine-1-phosphate (S1P). These effects may contribute to the increased risk of colorectal, breast, prostate, ovarian, and brain tumors reported for carriers of C282Y-HFE and explain why it is considered protective against Alzheimer’s disease. Collectively, previous reports (introduction of chapter 5) and my findings in this thesis indicated that C282Y-HFE markedly affected both cholesterol and sphingolipid metabolism and support the possibility that inhibition of both cholesterol synthesis/uptake as well as synthesis of S1P could be a potential therapeutic strategy for treatment of cancer in carriers of C282Y-HFE.

As indicated in figure 5.1, the findings in this thesis have made several important contributions to the understanding of how HFE variants affect disease processes. My findings have laid a foundation for several new fields of investigation that should be the subject of future research. The significance of key findings in this thesis as well as questions that logically stem from them are discussed below.

213

Figure 5.1: Summary of findings from this thesis

214 Is there a link between iron and cholesterol metabolism?

Numerous lines of research illustrate a connection between iron, cholesterol, and AD; however, the interaction of iron and cholesterol and its relevance to AD has not been specifically investigated. The interplay between iron and cholesterol is of significant relevance to understanding normal as well as abnormal physiology. This thesis explored the relationship between iron and cholesterol using cells expressing the clinically important H63D variant of HFE. I initially hypothesized that because H63D-HFE cells have increased iron content and iron is a required cofactor for a number of enzymes involved in cholesterol synthesis (eg. HMGCoAR and squalene synthase), increased amount of cholesterol would be synthesized by H63D-HFE cells. However, results obtained were the opposite; H63D-HFE cells expressed less HMGCoAR, had less total cholesterol and increased expression of CYP46A1. These observations were recapitulated in studies of the brains of H67D-HFE mice. Strikingly, a significant negative correlation was found between brain iron and cholesterol content. Our results indicating an inverse relationship between iron and cholesterol content in the H63D-HFE expressing cells and brains of mice expressing WT- or H67D-HFE are consistent with those reporting an association between increased brain (nigrostriatal) iron content and low serum cholesterol in patients with Parkinson’s disease (Du et al., 2012). Though it is believed that cholesterol content of serum is not truly reflective of the brain cholesterol, it is however important to note that in another study it was shown that patients with high serum iron and low serum cholesterol had the highest risk of PD (Powers et al., 2009), indicating the relevance of interplay between iron and cholesterol metabolism to neurodegenerative diseases.

An inverse relationship between iron and cholesterol content was also observed in several experimental studies. For example, Brunet et al. (1997) reported that treatment of rats with an iron-salicylate complex resulted in an elevation of iron, products of lipid peroxidation, and reduced total serum cholesterol. Treatment of rats with an iron dextran complex resulted in an elevation of iron but decrease in total serum cholesterol regardless of whether the rats were fed a normal diet or one high in cholesterol (Turbino-Ribeiro et al., 2003). Similar results had been found in earlier studies in which rabbits when treated

215 with an iron dextran complex and fed a normal diet were found to have reduction in serum cholesterol (Dabbagh et al., 1997). In fact, diet-induced iron overload resulted in reduced expression of HMGCoAR and CYP7A1 (Brunet et al., 1999). Though most of the mechanistic studies focused on effects of excess iron on plasma cholesterol, it is possible that excess iron has similar effects on cells in the brain. Our studies thus provide the first evidence for such an inverse relationship between iron and cholesterol content in the brain.

One possible explanation for these effects is that they are mediated by iron- induced oxidative stress (Dabbagh et al., 1994) and lipid peroxidation (Britton et al., 1987). Depletion of energy (ATP and NADPH) due to oxidative stress and lipid- peroxidation mediated membrane damage has been shown to cause disruption of lipid synthesis and transport (Kehrer, 1993). Our results along with previous studies described above support the hypothesis that the interplay between iron and cholesterol metabolism is complex, can affect the course of disease progression, and may involve additional mediators. One possible mediator in these interactions could be Aβ. The presence of the iron response element (IRE) in the promoter of the APP gene and the observation of an iron-concentration dependent increase in APP mRNA provide compelling evidence for the role of iron in APP expression and possibly Aβ generation (Bodovitz et al., 1995; Rogers and Lahiri, 2004; Rogers et al., 2002). Numerous reports have shown a link between elevated membrane cholesterol and increased processing of APP into Aβ (Canevari and Clark, 2007; Gellermann et al., 2005). In addition, a relationship between elevated sphingolipids, specifically ceramide or GM1 and increased Aβ accumulation has been reported (Hartmann et al., 2007; Zinser et al., 2007). Interestingly, there is also evidence that Aβ can, in turn, modulate cholesterol and sphingolipid metabolism by regulating the activities of key lipid enzymes (Grimm et al., 2012; Zinser et al., 2007). Specifically, Aβ has been shown to inhibit cholesterol synthesis by inhibiting HMGCoAR (Grimm et al., 2012), suggesting that a feedback mechanism exists between cellular cholesterol and Aβ content. Furthermore, GM1 has been demonstrated to serve as ‘seed’ for Aβ aggregation into toxic fibrils (Hayashi et al., 2004; Okada et al., 2007; Yanagisawa, 2005). Intriguingly, there is evidence that increase in GM1 can occur from reduction of cellular cholesterol content. In support of that, I found increased GM1 and

216 decreased cholesterol content in H63D-HFE, indicating a complex interplay between cholesterol, GM1, iron and Aβ.

Taking together findings from this thesis research plus studies of the role of APP metabolism in AD and its relationship to either iron or cholesterol as described above, I propose the following model to explain the mechanism by which the effects seen as a result of expression of H63D-HFE may lead to the symptoms associated with AD.

In my model, the increased cellular iron content induced by the expression of H63D-HFE elevates cholesterol content in the plasma membrane. The elevated iron and cholesterol then synergistically enhance Aβ accumulation and together induce oxidative stress and lipid peroxidation. As Aβ accumulates, it starts to inhibit HMGCoAR expression resulting in inhibition of cholesterol synthesis. As the cycle of increased iron, increased Aβ and inhibition of cholesterol synthesis by Aβ continues, the levels of oxidative radicals supersede the capacity of cell’s antioxidant capacity resulting in additional pathological processes including mitochondrial dysfunction, secretion of inflammatory cytokines and other toxic radicals, increase in tau phosphorylation, increase in intracellular Ca2+, and increase in GM1 and so on. In an effort to survive, cells may eventually adapt to a state where cholesterol content is reduced in an effort to limit Aβ production. Over time, possibly due to inability of cells to synthesize cholesterol and to cope with damages induced by oxidative radicals and other changes, myelin and eventually synapses disintegrate leading to cell death. Increased neuronal loss by this mechanism would then impair memory function culminating in AD.

With the exception of Aβ content, all of the pathological effects implicated in this model have been observed in either cells expressing H63D-HFE or in the brains of H67D-HFE mice. Cellular studies indicated no difference in the levels of total APP or Aβ between WT- and H63D-HFE, but increased sensitivity to Aβ treatment was observed in H63D-HFE cells (Mairuae et al., 2010). However, Aβ content in the brains of H67D- mice has not yet been evaluated and should be assessed in future. It is important to note that in our cell culture model, cells were stably transfected to express H63D-HFE. These cells have adapted in a manner that allows them to express H63D-HFE and survive, a system that may model age-related adaptation seen in humans. Thus, to further elucidate

217 the effects on iron on cholesterol synthesis, it is important to investigate these effects in cells that acutely express H63D-HFE i.e. tetracycline-inducible expression. It is possible that that acute expression of H63D will result in cholesterol elevation. In support of that argument, an increase in total brain cholesterol was observed in younger (6 month old) H67D-HFE mice relative to WT-mice (data not shown) while a significant reduction in brain cholesterol was observed in older animals (12-24 months). Therefore, further elucidation of age-related effects of the expression of H63D variant on APP metabolism in an in vivo model will shed light on the dynamics of HFE, APP, iron, and cholesterol metabolism, which are all key factors in AD pathogenesis.

As important as it is to study the impact of iron overload on lipid metabolism, it should also be of great interest to investigate the impact of high cholesterol on plasma and brain iron metabolism. Particularly, because of the established link between hypercholesterolemia and atherosclerosis and them being significant risk factors for AD and substantial evidence of disruption of iron metabolism and atherosclerosis, impact of high cholesterol on iron management should be studies. In fact, current evidence from studies in rabbit that were fed a high cholesterol diet showed increased iron accumulation in the brain (Ghribi et al., 2006; Ong and Halliwell, 2004; Ong et al., 2004) and elevation of lipid peroxidation products (De La Cruz et al., 2000; Gokkusu and Mostafazadeh, 2003). Though passage of dietary cholesterol into the brain is prohibited by the BBB, some metabolites, small molecules, and inflammatory cytokines produced by chronic hypercholesterolemia may still be able to cross from circulation into the BBB and eventually may even be able to affect the permeability of the BBB to other harmful substances. The observation of iron accumulation in the brain after long-term feeding on high- cholesterol diet provides support for this argument and raises the question whether and how hypercholesterolemia allows increased passage of iron into the brain parenchyma. The next logical question is then how would individuals with H63D-HFE be affected by these changes? We know that individuals with H63D-HFE have elevated iron levels in liver as well as in the brain (Adams et al., 2005; Bartzokis et al., 2010; Waheed et al., 1997). If these individuals develop hypercholesterolemia, they could have even more passage of iron into the brain, more oxidative stress and other disease associated consequences. Interestingly, it has been shown that individuals with ApoE4 allele have

218 high serum cholesterol and increased incidence of hypercholesterolemia. This provides another explanation for why individuals with both ApoE4 and H63D have increased risk and earlier onset of AD. This also explains why some epidemiological studies failed to see an interaction between ApoE4 and H63D as it could be that dietary choices and development of hypercholesterolemia are key determinants of the outcome. Further elucidation of this mechanism would provide explanation for why atherosclerosis and hypercholesterolemia- conditions that are far more common in western countries- may contribute to the development of AD. In relation to that, studies aiming to tease out the effects of diet (high-iron or high cholesterol, or both) on mice expressing WT-HFE (have normal iron; able to protect cells from iron overload) should be particularly important because results from those studies should enable us to determine the extent to which diet- induced changes affect iron and cholesterol metabolism in the brain.

How does H63D-HFE affect interactions between glial and neuronal cells?

Studies presented in this thesis along with previous reports have now established that H63D-HFE creates a permissive milieu for pathogenic processes (Ali-Rahmani et al., 2011; Hall et al., 2010; Hall et al., 2011; Lee et al., 2007; Liu et al., 2011; Mairuae et al., 2010; Mitchell et al., 2009a; Mitchell et al., 2009b). It would be important to test whether these effects can be recapitulated in neuronal cells isolated from H67D mice. Nonetheless, the impact of these changes may not be limited to neuronal cells alone. Therefore, the next question that stems from these findings is how would the effects of iron accumulation and resulting oxidative stress, elevation of extracellular glutamate, increased secretion of MCP-1 and 24S-HC on one cell type such as neurons impact other cells in the brain i.e. astrocytes, oligodendroglia, microglia, and endothelial cells and vice versa. The importance of this question is highlighted by the characteristic iron-deficient phenotype of macrophages in HH patients, a puzzle that has not yet been solved. It has been shown that although HH patients have a significant iron overload in their livers and other organs, their macrophages appear to be iron-depleted (Cardoso and de Sousa, 2003). Consistent with this observation, cellular studies have shown that the normal

219 function of WT-HFE is not only to manage iron uptake but to also prevent efflux of iron from cells into the extracellular space (Drakesmith et al., 2002). This function, however, is impaired in H63D-HFE cells resulting in more iron release from macrophages (Drakesmith et al., 2002). Considering that microglia in brain function similarly to macrophages in the circulation, a similar mechanism for regulating the CNS microenvironment might exist in the brain. If it does, it is likely that H63D-HFE expressing microglia would release more iron in the microenvironment of CNS which could be devastating to other cells especially oligodendrocytes, neurons, astrocytes and endothelial cells. This argument is supported by a number of studies demonstrating the deleterious effects of iron on cells in CNS. Zhang et al. showed that iron-loaded microglia increase the survival of oligodendrocytes by delivering iron in ferritin, which could contribute to the trophic effects on oligodendrocytes. However, lipopolysaccharide (LPS) activation of iron-loaded microglia attenuated the trophic effects and caused release of inflammatory cytokines from microglia (Zhang et al., 2007). In parallel, it has also been shown that microglial activation and secretion of inflammatory cytokines can decrease ferritin synthesis via modulatory effects of nitric oxide on the iron regulatory system (Chenais et al., 2002). It is logical to argue that a decrease in ferritin along with a pre-existing iron-overload and/or increased iron uptake would result in more free iron that could exacerbate oxidative damage that is often seen in neuronal and glial cells in neurological disorders (Hirsch and Faucheux, 1998; Rotig et al., 1997; Smith et al., 1998). So the question is what can cause microglial activation? Iron accumulation has been shown to influence glial activation either directly or through induction of other factors. One such factor is monocyte chemoattractant protein-1 (MCP-1) which is a chemokine (Rollins, 1997) that has been shown to cause glial activation (Vrotsos et al., 2009) as well as induction of pro-apoptotic genes and cell death (Zhou et al., 2006). In fact, previous studies have shown increased secretion of MCP-1 from cells expressing H63D-HFE (Mitchell et al., 2009a). Therefore, it is likely that iron accumulation within neurons can affect microglia and dysregulation of iron in microglia, in turn, influences the function of neuronal and glial cells, thereby creating a vicious cycle of events that potentiate stress and damage within the CNS. In such a scenario, endothelial cells, the gatekeepers of the BBB, could also be influenced by these changes. In fact, experimental

220 studies have shown that iron accumulation damages endothelial cells by induction of apoptotic pathways (Carlini et al., 2006). Therefore, maintaining iron homeostasis is essential for preserving endothelial function that is needed for maintaining integrity of the BBB and this has been shown to be compromised in several neurological disorders including AD.

Maintenance of the BBB is crucial for protecting the brain from the harmful effects of chemicals that can be present in the circulation and to keep a stable flux of ions into the brain for proper impulse generation and propagation by neurons. It is important to note that HFE, TfR and LDLR are expressed by the endothelial cells of the BBB. There is also evidence that iron can be taken up into the brain through the BBB. However, under normal physiological conditions no uptake of cholesterol from the circulation into the brain has been reported. Yet, based on the presence of LDLR on the BBB and evidence of inflammatory changes in the brains of animals fed a high- cholesterol diet, the possibility exists that under some pathological conditions circulating cholesterol or its metabolites or inflammatory cytokines may traverse the BBB. In this respect, understanding the effects of expression of H63D-HFE on endothelial cells would allow us to understand not only more of the dynamics of the CNS, but would also help us predict the impact of dietary components such as iron and cholesterol on brain metabolism and function.

The changes in iron and cholesterol induced by expression of H63D-HFE have major effects on the CNS and as such their effects on neuro-glial interactions should be further investigated. For such investigations, the H67D-HFE mouse model could continue to be valuable since findings presented in this thesis indicate that the animals expressed symptoms associated with human AD. It could also be used to study the effects of dietary manipulations such as high iron or high cholesterol diet, as well as therapeutic interventions such as statins and iron chelators on learning and memory and would provide valuable insight into gene-environment interactions which likely contribute to the pathogenesis of AD.

221 Because endothelial cells of brain microvasculature come in direct contact with the RBCs and exchange material i.e. ions, nutrients, etc, it was postulated that the characterization of RBCs may provide useful information about the brain metabolism. The changes in the membranes of RBC could reflect alterations in the brain and such alterations could be used as biomarkers to diagnose and monitor progression of neurological diseases. In fact, morphology of RBCs and the protein composition of RBC membranes from RBCs isolated from AD subjects have been shown to be altered (Mohanty et al., 2010; Sabolovic et al., 1997). Furthermore, oxidative stress associated changes in the phospholipid content were observed in RBC membranes isolated from RBCs from AD patients (Oma et al., 2012). These findings provide support for the notion that RBC lipid dynamics could be important determinants in neurological diseases. Accordingly, studies presented in this thesis showed that neuronal cells expressing H63D-HFE had a decrease in cholesterol that was associated with increased GM1 relative to the WT-HFE controls (Chapter 2). Moreover, RBC membranes from autistic children also showed reduced cholesterol and increased GM1 relative to healthy controls (Appendix). Based on the similar biochemical phenotype observed in neuronal cells expressing H63D-HFE and RBCs from autistic children, a logical and important future avenue of investigation would be to analyze lipid and protein composition of RBC membranes from individuals with H63D-HFE, those that are healthy as well as those with neurological disorders. Given that it is much easier to obtain a blood sample from individuals, identification of biomarkers in the blood would be an ideal approach. If indeed a biochemical profile is identified in RBCs from carriers of HFE variants (control, MCI, AD) that is associated with disease pathology, we could evaluate whether it could be used as biomarker to study disease progression. Identification of such a biomarker would significantly aid in future investigations and possibly treatment. Furthermore, our finding that decreased brain cholesterol was associated with disruption of myelin and synapse management as well as cognitive function, provides compelling evidence to evaluate whether H63D-HFE is associated with increased risk of developing autism spectrum disorders.

222 Potential biomarkers for AD progression

It has been well established that brain lipid content is crucial for maintenance of normal brain function. However, the brain is a relatively analytically inaccessible compartment of the human body, especially when it comes to analyzing its lipid content in living individuals. Due to the magnetic properties of iron, brain iron status can be accessed, at least indirectly, by MRI (Bartzokis et al., 1999). However, no such tool is available to monitor the status of brain cholesterol. Owing to the high lipid content of myelin, certain MRI techniques that measure myelination by way of quantifying white matter tracts may provide insight into lipid status, but no direct conclusion can be made about brain lipid content. Therefore, there has been tremendous interest in identifying biomarkers associated with onset, progression, and response to treatments for brain disorders. With respect to brain cholesterol homeostasis, one candidate that could be of pathophysiological relevance is 24S-hydroxycholesterol (24S-HC). Because almost all 24S-HC found in plasma and CSF is known to originate from the brain due to the action of a brain specific enzyme, CYP46A1, it has been hypothesized that 24S-HC can be used as an indicator of brain cholesterol turnover (Bjorkhem et al., 1998). This led to a number of studies, the results of which have shown an elevation of 24-HC either in the CSF or in the plasma of patients suffering from traumatic brain injury or neurodegenerative diseases (REF). In particular, it was found that during the early stages of AD or vascular dementia (Lutjohann et al., 2000) and active periods of multiple sclerosis (MS), high levels of 24S-HC were present in either the CSF or circulation (Leoni et al., 2002), possibly due to ongoing demyelination (Bjorkhem and Meaney, 2004). However, in late stage AD patients CSF levels of 24S-HC were found to be reduced (Bretillon et al., 2000; Leoni et al., 2002; Leoni et al., 2006; Papassotiropoulos et al., 2002); they were also reduced in the brains of patients who died from AD (Heverin et al., 2004). The reduction probably reflected the extensive neuronal loss seen in late stages of AD. Consistent with these findings, I found an elevation in expression of the neuron-specific enzyme CYP46A1 that catalyzes conversion of brain cholesterol to 24S-HC in both human neuroblastoma cells expressing H63D-HFE and in the brains of H67D-HFE mice. Though I did not measure the amount of 24S-HC in this study, based on the concomitant decrease in cellular and total brain cholesterol it is likely that H63D/H67D expression

223 would be associated with higher levels of 24S-HC in serum or CSF or both providing further support for the potential use of 24S-HC for tracing the progression of AD and associated changes in the brain cholesterol content. It would be useful to confirm this hypothesis using the H67D mouse model. These findings could then provide a rationale for measuring the levels of 24S-HC in the plasma or CSF of individuals carrying H63D- HFE. If a relationship could be established between levels of 24S-HC and incidence or severity of AD, it could be used as a biomarker.

It is also possible that brain volume as determined by MRI could be used as a biomarker. This hypothesis is based on the early reports of elevated brain iron in the carriers of H63D (Bartzokis et al., 2010), as well as on the finding of this thesis that reduced brain volume in aged H67D-HFE expressing mice, a finding in agreement with the concomitant neurodegeneration shown by immunohistochemistry. The decline in brain volume correlated significantly with brain cholesterol, brain iron content, and measures of learning and memory. Therefore, these findings suggest that the progression of memory decline could be monitored by using volumetric MRI measurements. These tools could in turn be useful for accessing the efficacy of treatments as well in patients. In summary, findings in this thesis have identified two biomarkers that could be beneficial for monitoring disease progression and possibly treatment efficacy in the carriers of H63D-HFE.

Is it iron accumulation or expression of the HFE variants that is the culprit?

H63D and C282Y are two most common variants of HFE and expression of either is known to result in iron overload. Additionally, a number of other cellular processes are also altered in cells expressing these variants. This raises the question whether observed changes are resultant of iron accumulation or occur due to the gain-of-function of the mutation. Epidemiological studies have indicated H63D-HFE as a risk factor for AD and C282Y-HFE for certain cancers. Results presented in this thesis (summarized in Table 5.1) corroborate those results. The fact that both variants of HFE, H63D and C282Y, are associated with intracellular iron accumulation, yet result in opposing phenotypes, H63D-

224 HFE promoting apoptosis and C282Y-HFE promoting cell survival, indicates that these mutations by themselves could be playing a role independent/ in addition to their effect on cellular iron status. Supporting this argument are the findings that cellular localization of both variants is different. Both WT- and H63D-HFE localize in the plasma membrane, while C282Y-HFE is retained in ER/golgi. This striking difference is important because disruption of cellular localization could reflect alteration in protein-protein interactions and cell signaling. Indeed, I found that while WT-HFE and H63D-HFE were localized in fractions containing lipid rafts, C282Y-HFE was associated with soluble/ non-raft domains. Because lipid rafts provide a platform for protein- protein interactions for activation or inhibition of cell signaling pathways, an intriguing possibility is that C282Y-HFE may not interact with a protein with which WT-HFE does to maintain normal homeostasis. Here it is important to note that although H63D-HFE localizes in the plasma membrane, possibly in regions just outside of rafts, its interaction with TfR is altered such that it loses the ability to prevent Tf binding to TfR and iron uptake. Therefore, interactions of both of these mutants with other proteins may be affected. Currently, other binding partners of WT and mutant HFE proteins are not known. Further investigation in this area is needed to elucidate the role of HFE in cell signaling. There may even be a direct, yet unidentified, role for HFE in lipid metabolism that could be altered as a result of mutation. Nonetheless, current evidence indicates that mutation as well as resulting iron overload could contribute to the observed effects.

225 Table 5.1: Comparison of in vivo and in vitro findings of H63D- and C282Y-HFE

H63D C282Y

Iron ↑ ↑ Oxidative stress ↑ ↑ Endoplasmic Cellular location Plasma Membrane Reticulum/ Golgi Cellular cholesterol ↓↓ ↑↑

GM1 ↑↑ ↓↓ Yes (marginally Lipid Raft No localized) ↓Cell Proliferation ↑Cell Proliferation Cellular phenotype Apototic

Cancer; Disease risk AD Protective for AD ↑ in early life ------Plasma cholesterol ↓ in later life

In Mice

Plasma cholesterol ↓ in Humans

226 Another potential area for future investigation would be to identify proteins that serve as molecular switches in the cell and whose expression and/ or activity are directly regulated by cellular iron content. This would help us dissect the molecular mechanism underlying a wide array of processes and diseases such as AD and cancer in which iron is implicated. One possible candidate is peptidyl-prolyl cis-trans isomerase (Pin1). It is known to play a key role in regulation of signal transduction pathways (Wulf et al., 2005), in particular, cellular proliferation (Lu et al., 1996). Phosphorylation of Pin1 determines its activity i.e. more phosphorylation is indicative of decreased activity. Pin1 modulates the folding, activity, and stability of target proteins via isomerization. The findings that Pin1 is downregulated in degenerating neurons from Alzheimer disease patients (Liou et al., 2003) but its expression/ activity is elevated in many cancers including those of the breast, prostate, brain, lung, and colon (Ayala et al., 2003; Bao et al., 2004; Ryo et al., 2001; Wulf et al., 2001), provide support for the argument that Pin1 can serve as a molecular switch in the cell. Interestingly, varying concentrations of iron were shown to alter Pin1 phosphorylation and activity in WT- and H63D-HFE cells, indicating that iron can modulate Pin1 activity (Hall et al., 2010). Therefore, it would be valuable to investigate Pin 1 expression and activity in cells expressing C282Y-HFE. If higher expression and/ activity is observed in C282Y-HFE cells, it would provide insight into the mechanism underlying the cancerous phenotype associated with expression of C282Y-HFE. Additionally, this knowledge should provide an insight into how iron overload contributes to both neurodegenerative diseases as well as cancer.

Therapeutic implications: Is statin therapy appropriate for AD?

The treatments of neurological disorders often reflect a “one size fits all” approach, perhaps because there is insufficient research dedicated towards understanding the interface different disease pathways and the environment and genetic risk factors. Because AD is a multifactorial disease with evidence for the involvement of multiple pathways and genetic mutations in genes with diverse function, it would be beneficial to identify subpopulations responsive to specific treatments based on their genotype. Identification of genetic and environmental factors, knowledge about their interactions,

227 and resulting pathological markers will allow us to improve therapeutic outcomes. Therefore, I attempted to ascertain whether there is crosstalk between cholesterol and iron metabolism and how it may contribute to pathogenesis of diseases. Results presented in this thesis could have major implications on therapeutic strategies for AD treatment. I predict that a long-range outcome would be stratification of medical interventions based on HFE genotype. For example, I propose that H63D-HFE positive patients with AD will differ from those with WT-HFE in their response to statin therapy based on our observation that cells with the H63D-HFE allele have lower baseline levels of total cholesterol and exhibit slower cellular growth relative to cells expressing WT-HFE. Our finding that treatment of H63D-HFE cells with a statin that can cross the BBB resulted in decreased cell survival supports this proposal. These data imply that lowering CNS cholesterol would be deleterious to neuronal function, and more so in the carriers of H63D-HFE. Further support for this proposal comes from a recent clinical trial that showed a reduction in right hippocampal volume after 1 year of atorvastatin therapy in AD patients (Sparks et al., 2008). These findings suggest that in addition to lowering plasma cholesterol levels in AD patients, statin treatment may have reduced brain cholesterol possibly resulting in neuronal loss and atrophy of certain brain regions. Thus, care should be taken when administering to patients statins that can cross the blood-brain barrier and directly inhibit cholesterol synthesis in brain cells, as a possible consequence is the perturbation of the numerous physiological functions for which cholesterol is needed including synaptic transmission and learning and memory. Instead, efforts should be made to develop drugs that can prevent neuronal loss of cholesterol while reducing its levels in the circulation.

228 References

Ali-Rahmani, F., J.A. Hengst, J.R. Connor, and C.L. Schengrund. 2011. Effect of HFE Variants on Sphingolipid Expression by SH-SY5Y Human Neuroblastoma Cells. Neurochem Res.

Ayala, G., D.G. Wang, G. Wulf, A. Frolov, R. Li, J. Sowadski, T.M. Wheeler, K.P. Lu, and L. Bao. 2003. The prolyl isomerase Pin1 is a novel prognostic marker in human prostate cancer. Cancer Research. 63:6244-6251.

Bao, L., A. Kimzey, G. Sauter, J.M. Sowadski, K.P. Lu, and D.G. Wang. 2004. Prevalent overexpression of prolyl isomerase Pin1 in human cancers. American Journal of Pathology. 164:1727-1737.

Bartzokis, G., J.L. Cummings, C.H. Markham, P.Z. Marmarelis, L.J. Treciokas, T.A. Tishler, S.R. Marder, and J. Mintz. 1999. MRI evaluation of brain iron in earlier- and later-onset Parkinson's disease and normal subjects. Magn Reson Imaging. 17:213-22.

Bartzokis, G., P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P. Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely, and J.R. Connor. 2010. Prevalent iron metabolism gene variants associated with increased brain ferritin iron in healthy older men. J Alzheimers Dis. 20:333-41.

229 Bjorkhem, I., and S. Meaney. 2004. Brain cholesterol: Long secret life behind a barrier. Arteriosclerosis Thrombosis and Vascular Biology. 24:806-815.

Bodovitz, S., M.T. Falduto, D.E. Frail, and W.L. Klein. 1995. Iron Levels Modulate Alpha-Secretase Cleavage of Amyloid Precursor Protein. Journal of Neurochemistry. 64:307-315.

Bretillon, L., D. Lutjohann, L. Stahle, T. Widhe, L. Bindl, G. Eggertsen, U. Diczfalusy, and I. Bjorkhem. 2000. Plasma levels of 24S-hydroxycholesterol reflect the balance between cerebral production and hepatic metabolism and are inversely related to body surface. J Lipid Res. 41:840-5.

Britton, R.S., B.R. Bacon, and R.O. Recknagel. 1987. Lipid-Peroxidation and Associated Hepatic Organelle Dysfunction in Iron Overload. Chemistry and Physics of Lipids. 45:207-239.

Canevari, L., and J.B. Clark. 2007. Alzheimer's disease and cholesterol: the fat connection. Neurochem Res. 32:739-50.

Cardoso, C.S., and M. de Sousa. 2003. HFE, the MHC and hemochromatosis: Paradigm for an extended function for MHC class I. Tissue Antigens. 61:263-275.

Carlini, R.G., E. Alonzo, E. Bellorin-Font, and J.R. Weisinger. 2006. Apoptotic stress pathway activation mediated by iron on endothelial cells in vitro. Nephrology Dialysis Transplantation. 21:3055-3061.

Chenais, B., H. Morjani, and J.C. Drapier. 2002. Impact of endogenous nitric oxide on microglial cell energy metabolism and labile iron pool. J Neurochem. 81:615-23.

230 Dabbagh, A.J., T. Mannion, S.M. Lynch, and B. Frei. 1994. The Effect of Iron Overload on Rat Plasma and Liver Oxidant Status in-Vivo. Biochemical Journal. 300:799- 803.

De La Cruz, J.P., L. Quintero, M.A. Villalobos, and F. Sanchez de la Cuesta. 2000. Lipid peroxidation and glutathione system in hyperlipemic rabbits: influence of olive oil administration. Biochim Biophys Acta. 1485:36-44.

Drakesmith, H., E. Sweetland, L. Schimanski, J. Edwards, D. Cowley, M. Ashraf, J. Bastin, and A.R.M. Townsend. 2002. The hemochromatosis protein HFE inhibits iron export from macrophages. Proceedings of the National Academy of Sciences of the United States of America. 99:15602-15607.

Du, G.W., M.M. Lewis, M.L. Shaffer, H.L. Chen, Q.X. Yang, R.B. Mailman, and X.M. Huang. 2012. Serum Cholesterol and Nigrostriatal R2*Values in Parkinson's Disease. PLoS One. 7.

Gellermann, G.P., T.R. Appel, A. Tannert, A. Radestock, P. Hortschansky, V. Schroeckh, C. Leisner, T. Lutkepohl, S. Shtrasburg, C. Rocken, M. Pras, R.P. Linke, S. Diekmann, and M. Fandrich. 2005. Raft lipids as common components of human extracellular amyloid fibrils. Proc Natl Acad Sci U S A. 102:6297-302.

Ghribi, O., M.Y. Golovko, B. Larsen, M. Schrag, and E.J. Murphy. 2006. Deposition of iron and beta-amyloid plaques is associated with cortical cellular damage in rabbits fed with long-term cholesterol-enriched diets. J Neurochem. 99:438-49.

Gokkusu, C., and T. Mostafazadeh. 2003. Changes of oxidative stress in various tissues by long-term administration of vitamin E in hypercholesterolemic rats. Clin Chim Acta. 328:155-61.

231 Grimm, M.O.W., T.L. Rothhaar, and T. Hartmann. 2012. The role of APP proteolytic processing in lipid metabolism. Experimental Brain Research. 217:365-375.

Hall, E.C., 2nd, S.Y. Lee, Z. Simmons, E.B. Neely, W. Nandar, and J.R. Connor. 2010. Prolyl-peptidyl isomerase, Pin1, phosphorylation is compromised in association with the expression of the HFE polymorphic allele, H63D. Biochim Biophys Acta. 1802:389-95.

Hartmann, T., J. Kuchenbecker, and M.O.W. Grimm. 2007. Alzheimer's disease: the lipid connection. Journal of Neurochemistry. 103:159-170.

Heverin, M., N. Bogdanovic, D. Lutjohann, T. Bayer, I. Pikuleva, L. Bretillon, U. Diczfalusy, B. Winblad, and I. Bjorkhem. 2004. Changes in the levels of cerebral and extracerebral sterols in the brain of patients with Alzheimer's disease. Journal of Lipid Research. 45:186-193.

Hirsch, E.C., and B.A. Faucheux. 1998. Iron metabolism and Parkinson's disease. Movement Disorders. 13:39-45.

Kehrer, J.P. 1993. Free-Radicals as Mediators of Tissue-Injury and Disease. Critical Reviews in Toxicology. 23:21-48.

232 Leoni, V., T. Masterman, U. Diczfalusy, G. De Luca, J. Hillert, and I. Bjorkhem. 2002. Changes in human plasma levels of the brain specific oxysterol 24S- hydroxycholesterol during progression of multiple sclerosis. Neurosci Lett. 331:163-6.

Leoni, V., M. Shafaati, A. Salomon, M. Kivipelto, I. Bjorkhem, and L.O. Wahlund. 2006. Are the CSF levels of 24S-hydroxycholesterol a sensitive biomarker for mild cognitive impairment? Neurosci Lett. 397:83-7.

Liou, Y.C., A. Sun, A. Ryo, X.Z. Zhou, Z.X. Yu, H.K. Huang, T. Uchida, R. Bronson, G.Y. Bing, X.J. Li, T. Hunter, and K.P. Lu. 2003. Role of the prolyl isomerase Pin1 in protecting against age-dependent neurodegeneration. Nature. 424:556- 561.

Lu, K.P., S.D. Hanes, and T. Hunter. 1996. A human peptidyl-prolyl isomerase essential for regulation of mitosis. Nature. 380:544-547.

Mitchell, R.M., S.Y. Lee, W.T. Randazzo, Z. Simmons, and J.R. Connor. 2009a. Influence of HFE variants and cellular iron on monocyte chemoattractant protein- 1. J Neuroinflammation. 6:6.

233 Mitchell, R.M., S.Y. Lee, Z. Simmons, and J.R. Connor. 2009b. HFE polymorphisms affect cellular glutamate regulation. Neurobiol Aging.

Mohanty, J.G., H.D. Shukla, J.D. Williamson, L.J. Launer, S. Saxena, and J.M. Rifkind. 2010. Alterations in the red blood cell membrane proteome in alzheimer's subjects reflect disease-related changes and provide insight into altered cell morphology. Proteome Science. 8.

Ong, W.Y., and B. Halliwell. 2004. Iron, atherosclerosis, and neurodegeneration: a key role for cholesterol in promoting iron-dependent oxidative damage? Ann N Y Acad Sci. 1012:51-64.

Ong, W.Y., B. Tan, N. Pan, A. Jenner, M. Whiteman, C.N. Ong, F. Watt, and B. Halliwell. 2004. Increased iron staining in the cerebral cortex of cholesterol fed rabbits. Mech Ageing Dev. 125:305-13.

Powers, K.M., T. Smith-Weller, G.M. Franklin, W.T. Longstreth, P.D. Swanson, and H. Checkoway. 2009. Dietary fats, cholesterol and iron as risk factors for Parkinson's disease. Parkinsonism & Related Disorders. 15:47-52.

Rogers, J.T., and D.K. Lahiri. 2004. Metal and inflammatory targets for Alzheimer's disease. Curr Drug Targets. 5:535-51.

234 Rogers, J.T., J.D. Randall, C.M. Cahill, P.S. Eder, X.D. Huang, H. Gunshin, L. Leiter, J. McPhee, S.S. Sarang, T. Utsuki, N.H. Greig, D.K. Lahiri, R.E. Tanzi, A.I. Bush, T. Giordano, and S.R. Gullans. 2002. An iron-responsive element type II in the 5 '-untranslated region of the Alzheimer's amyloid precursor protein transcript. Journal of Biological Chemistry. 277:45518-45528.

Rollins, B.J. 1997. Chemokines. Blood. 90:909-928.

Rotig, A., P. deLonlay, D. Chretien, F. Foury, M. Koenig, D. Sidi, A. Munnich, and P. Rustin. 1997. Aconitase and mitochondrial iron-sulphur protein deficiency in Friedreich ataxia. Nature Genetics. 17:215-217.

Ryo, A., M. Nakamura, G. Wulf, Y.C. Liou, and K.P. Lu. 2001. Pin 1 regulates turnover and subcellular localization of beta-catenin by inhibiting its interaction with APC. Nature Cell Biology. 3:793-801.

Smith, M.A., K. Wehr, P.L.R. Harris, S.L. Siedlak, J.R. Connor, and G. Perry. 1998. Abnormal localization of iron regulatory protein in Alzheimer's disease. Brain Research. 788:232-236.

Sparks, D.L., S.K. Lemieux, M.W. Haut, L.C. Baxter, S.C. Johnson, L.M. Sparks, H. Sampath, J.E. Lopez, M.H. Sabbagh, and D.J. Connor. 2008. Hippocampal volume change in the Alzheimer Disease Cholesterol-Lowering Treatment trial. Cleve Clin J Med. 75 Suppl 2:S87-93.

235 Vrotsos, E.G., P.E. Kolattukudy, and K. Sugaya. 2009. MCP-1 involvement in glial differentiation of neuroprogenitor cells through APP signaling. Brain Research Bulletin. 79:97-103.

Wulf, G., G. Finn, F. Suizu, and K.P. Lu. 2005. Phosphorylation-specific prolyl isomerization: is there an underlying theme? Nature Cell Biology. 7:435-441.

Wulf, G.M., A. Ryo, G.G. Wulf, S.W. Lee, T.H. Niu, V. Petkova, and K.P. Lu. 2001. Pin1 is overexpressed in breast cancer and cooperates with Ras signaling in increasing the transcriptional activity of c-Jun towards cyclin D1. Embo Journal. 20:3459-3472.

Yanagisawa, K. 2005. GM1 ganglioside and the seeding of amyloid in Alzheimer's disease: endogenous seed for Alzheimer amyloid. Neuroscientist. 11:250-60.

Zhang, X., N. Surguladze, B. Slagle-Webb, A. Cozzi, and J.R. Connor. 2007. Cellular iron status influences the functional relationship between microglia and oligodendrocytes. American Journal of Hematology. 82:573-573.

Zhou, L.M., A. Azfer, J.L. Niu, S. Graham, M. Choudhury, F.M. Adamski, C. Younce, P.F. Binkley, and P.E. Kolattukudy. 2006. Monocyte chemoattractant protein-1 induces a novel transcription factor that causes cardiac myocyte apoptosis and ventricular dysfunction. Circulation Research. 98:1177-1185.

Zinser, E.G., T. Hartmann, and M.O.W. Grimm. 2007. Amyloid beta-protein and lipid metabolism. Biochimica Et Biophysica Acta-Biomembranes. 1768:1991-2001.

236 Appendix

Cholesterol, GM1, and Autism

Abstract

Disruption of cholesterol metabolism has been hypothesized to contribute to dementia, possibly due to its role in maintaining membrane fluidity as well as the integrity of lipid rafts. Previously, we reported an apparent inverse relationship between membrane cholesterol levels and those of GM1, another lipid that can be found in rafts. This paper describes the observation that red blood cell (RBC) membranes isolated from blood drawn from children diagnosed with autism have on the average significantly less cholesterol and significantly more GM1 than RBC membranes isolated from blood obtained from control children. While cholesterol in the circulation does not cross the blood brain barrier, a generalized defect in its synthesis could affect its concentration in the central nervous system and that, coupled with a change in ganglioside expression, could contribute to development of the behaviors associated with autism.

Introduction

Autism is defined by an individual’s behavior and is one of several pervasive developmental disorders. According to NINDS, experts estimate that three to six children out of every 1,000 will have autism with males four times more likely than females to be affected [1]. Children with autism have impaired social interaction, abnormal communication skills, and repetitive behaviors or interests. It is these behaviors that are used to identify children with autism [usually as set forth in 2]. Because the behaviors may be subtle, or overlap with typical development in infants and toddlers, it is often not diagnosed until a child is three years of age or older. Delay in diagnosis is problematic as

237 the earlier autism is identified and treatment started, the more likely that improvement will be noticed.

Analysis of a number of factors including genetic, metabolic and environmental has failed to identify the primary cause for autism. Studies focused on lipids present in the blood of affected children have given mixed results. A report by Bu et al [3] on the fatty acids present in erythrocyte membranes obtained from autistic subjects and controls did not support the hypothesis that a defect in fatty acid metabolism was the underlying cause of autism. These observations contradicted those in an earlier paper that reported the fatty acid composition of erythrocyte phospholipids obtained from a single patient contained more highly unsaturated fatty acids than controls [4]. The apparent difference in results emphasizes the need to study multiple samples to determine whether there is a consistent difference regardless of the conditions under which samples are obtained.

Evidence for alterations in cholesterol metabolism in at least a subset of children with autism spectrum disorders is accumulating. Kelley [5] found that patients with Smith-Lemli-Opitz syndrome (SLOS) had elevated levels of 7-dehydrocholesterol, the immediate precursor in the synthesis of cholesterol, due to a deficiency in 7- dehydrocholesterol reductase [6], the enzyme that catalyzes reduction of 7- dehydrocholesterol to cholesterol. Because about 50% of SLOS patients meet the criteria for autism the finding of elevated 7-dehydrocholesterol was hypothesized to serve as a possible model for understanding the genetic causes of autism and the role of cholesterol in autism spectrum disorder [7]. Research by Tierney et al [8] in which serum samples from 100 children with autism were studied indicated that 19 of the samples had cholesterol levels below the 5th percentile for children over 2 yrs of age. If the level of free cholesterol found in cell membranes is reduced that could affect development and possibly formation of lipid rafts – areas of cell membranes known to be enriched in cholesterol, sphingolipids, and signal transduction molecules. Cholesterol is essential for myelination [9] and research has indicated that low levels of cholesterol may contribute to development of neuronal problems [10].

In previous work we observed that acute disruption of cholesterol using either methyl-ß-

238 filipin (aggregates membrane cholesterol [12]) was accompanied by changes in axonogenesis [13], enhanced cleavage of SNAP-25 by botulinum neurotoxin serotype A [14], and enhanced exo- and endo-cytosis [15] by N2a murine neuroblastoma cells. Cholesterol has also been shown to affect the oxytocin receptor [16], which when absent in male mice results in pronounced deficits in social recognition, obesity, and lack of body temperature control [17], and the G-protein coupling of the serotonin1A receptor which functions during development to establish anxiety-like behavior [18].

found to affect expression of gangliosides GM1 and GD1a, and enhance association of GM1 with the nucleus [13]. Interestingly, increased levels of GM1 were found in cerebral spinal fluid from children with autism spectrum disorders [19] and increased serum levels of anti-GM1 auto-antibodies were present in children diagnosed with autism [20]. Combined, these observations led us to look at cholesterol and GM1 present in membranes isolated from red blood cells to ascertain whether they differed in samples prepared from blood obtained from children with autism relative to levels present in samples obtained from controls.

Methods

Approval for study of not more than 20 children diagnosed as having autism and 20 controls was obtained from the Human Subjects Protection Office (IRB Protocol # 30838EP) prior to initiation of this study. Children enrolled in the study were evaluated for symptoms of autism using a two-step procedure. Initially each potential enrollee was evaluated in the pediatric autism clinic by an experienced developmental pediatrician. Individuals that met DSM-IV-R criteria to have autism, had their diagnosis confirmed using an interview specialized for research use, the Autism Diagnostic Interview-Revised (ADI-R [2]). Scores obtained from this interview were used to grade severity (see Table 1, AD score). Comparison children (controls) were evaluated using the Modified Checklist for Autism in Toddlers (M-CHAT) for children 3-5 years 11 months and the Social Responsiveness Scale for children 6 -9 yrs completed by a parent. These standardized questionnaires were utilized in order to exclude children with symptoms of autism or other developmental disorders from being included in the control group. The

239 children in both groups ranged in age from 3 to 8 years. Sixteen children (1female, 15 males) with autism and 20 comparison children (12 females, 8 males) were analyzed. The average age for the cohort with autism was 5.13 yrs, while it was 6.0 yrs for the control group. Participants’ age, sex and AD score, when applicable, are given in Table 1.

Three ml of blood, drawn in tubes with zinc free stoppers/no anticoagulant, were centrifuged to remove serum and the pellets recovered taken for analysis. Erythrocytes, dispersed by addition of an equal volume of balanced salt solution (pH7.6) to each pellet, were isolated by centrifugation on Ficoll-Paque™ Premium (GE Healthcare, Piscataway, NJ) at 400xG for 30 min at room temperature. Erythrocytes, recovered at the bottom of the tube, were washed (20mM Tris-HCl containing 0.13M NaCl, pH 7.4) prior to lysis in cold 5mM sodium phosphate, pH8.0. Membranes were recovered by centrifugation (14,000xG, 4˚C) and washed until hemoglobin free. Protein concentration was determined using the Bio-Rad protein assay (Bio-Rad, Hercules, CA) with bovine serum albumin as the standard. Samples were then stored at -80˚C until used.

Cholesterol was extracted from 0.1mg of isolated membrane protein using chloroform: isopropanol:NP‐40 (7:11:0.1 by vol). The amount of cholesterol recovered in the chloroform fraction was determined using the BioVision cholesterol assay (Mountain View, CA) according to the manufacturer’s instructions. Since membranes contain cholesterol and not cholesterol esters, cholesterol esterase was not used to catalyze hydrolysis of fatty acid from cholesterol prior to analysis. In this assay cholesterol oxidase is used to oxidize the hydroxyl group on cholesterol to the ketone with the concomitant production of hydrogen peroxide. The hydrogen peroxide reacts with the probe and the color produced visualized using a wavelegth of 570nm. Known concentrations of cholesterol were used as standards.

Presence of GM1 in RBC membranes was monitored using slot/dot blot technology [21, 22]. Two µg of RBC membrane protein was blotted (Minifold® II Slot- Blot System, Whatman, VWR International) onto a nitrocellulose membrane according to the manufacturer’s directions. Each sample was run in duplicate. Membranes were blocked for 1hr at room temperature using 5% nonfat dried milk in phosphate buffered saline prior to probing overnight at 4˚C with horse radish peroxidase conjugated binding

240 subunit of cholera toxin (CTxB, List Biological Laboratories, Campbell, CA) in PBS containing 2% nonfat dried milk. Bands were visualized using Super Signal West Femto Maximum Sensitivity Substrate™ followed by exposure in a Fuji Film LAS 3000. Multigauge software was used to calculate intensities. GM1 in erythrocytes was identified chemically by Kundu et al [23].

Results

Gender, age, and ADI-R score are given in Table 1 for children from whom samples were obtained and used for analysis. Significant differences in cholesterol content were not observed between samples isolated from RBCs from either control males or females, nor were age-related differences apparent in the results from samples obtained from control children ages 4 through 8 yrs. This was important to assess because of differences between the affected and control groups. There were significantly more girls in the comparison group and the average age was somewhat older (6.0 yrs for controls; 5.13 yrs for affected). Based on previous observations indicating cholesterol levels in blood could be altered in children with autism [5, 8] and because cholesterol is not only a component of cell membranes but is an essential component of erythrocyte lipid rafts [24], its level in membranes isolated from RBCs was determined. A significant decrease (p = 0.012, 2-tailed t-test) was seen in the average amount of cholesterol in membranes isolated from RBCs from children diagnosed with autism compared to those isolated from controls (see Figure 1A). No correlation was seen between scores obtained using the ADI-R and cholesterol levels (Figure 1B).

This observation led us to determine the relative amounts of GM1, a known component of RBC lipid rafts [25, 26] in membranes of RBCs isolated from both children diagnosed with autism and controls (* next to the age of children indicated in Table 1 identifies samples used in these assays). While the sample size was limited, the results from slot blot analyses (Figure 2) indicate that there was on the average significantly more (p = 0.019, unpaired, 2-tailed t-test) GM1 in the membranes of RBCs isolated from children with autism than controls. Analysis of the adherence of the binding subunit of cholera toxin to blots of proteins separated by SDS-PAGE indicated the presence of a single positive band that ran at the front of the blot. The observed decrease

241 in cholesterol coupled with the increase in expression of GM1 in RBC membranes from children diagnosed with autism shown in Figure 3, is in agreement with our previous observations that loss of membrane cholesterol was accompanied by an increase in GM1 [13, 27].

Discussion

Increasingly, autism is hypothesized to be a whole body disorder [eg 28] or to involve metabolic pathways expressed throughout the entire body, which, if correct, suggests that it should be possible to detect it using non-neural samples. The concept that a “neural” problem may reflect a generalized defect of a functional element has been examined in people with depression where researchers postulated that they could evaluate treatment for depression by looking at the association of Gs with lipid rafts in platelets [29] and erythrocytes [30]. Those observations, plus previous studies of blood from autistic children [eg 8] provided the rationale for monitoring cholesterol and GM1 in RBC membranes.

The lack of correlation between scores obtained using the ADI-R and cholesterol levels (Figure 1B) was evident not only for the total score but for subscales that measured level of impairment of social interaction, limitation of communication and presence of repetitive and stereotypic behaviors as well. While a relationship would add to the potential significance of the current observations, the Autism Diagnostic Interview- Revised was devised to provide a clear diagnosis of classic autism for research purposes. In initial investigations like this one, it is critical to study a sample of children who clearly have autism as it is currently recognized. Because the ADI-R was not designed to quantify symptoms in the domains it samples (social atypia, communication dysfunction and repetitive, stereotypic and restrictive behaviors) in future studies it will be important to use instruments to evaluate the range of autism-associated symptoms to allow for potential correlation of behavioral phenotype with lipid profile.

While the results indicate a significant average decrease in cholesterol content of RBC membranes isolated from children with autism it can be seen that some of the values were in the normal range, in agreement with previous observations [8] indicating

242 that only 19% of children with autism had significantly lower serum cholesterol levels than controls. Evidence that alterations in cholesterol metabolism may occur in children with autism is provided by the observation that expression of ApoB-100, a protein needed for transport of very low density lipoproteins (VLDL) from the liver to the circulation as well as for uptake of low density lipoproteins (LDL) from the circulation, was lower in children with autism compared to controls (31). Interestingly, in a case controlled study of 29 autistic boys and 29 similar male controls [32] no significant difference in the levels of total serum cholesterol and LDL-cholesterol were observed while a significant reduction in high density lipoprotein-cholesterol and a significant increase in triacylglycerides (TAGs) were found. These observations might be explained by failure of cells to appropriately metabolize TAG-rich VLDLs produced by the liver to LDLs as well as by a failure of cells to take up the LDLs that are produced. The latter may reflect altered expression of LDL receptors on the cell surface. Their finding of elevated LDL/HDL ratios in the autistic cohort supports the observation that adults with autism spectrum disorder are more apt to be diagnosed with hyperlipidemia and as a result are considered to be at risk for developing coronary heart disease [33]. While these observations might appear to contradict the hypothesis that there may be a role for cholesterol in the treatment of autism [34], that recommendation was made for autistic children found to be deficient in 7-dehydrocholesterol reductase. Overall, accumulating evidence supports the fact that a significant number of children diagnosed with autism spectrum disorder have alterations in cell cholesterol content/metabolism.

Cholesterol accounts for between 25 and 50% of lipid found in plasma membranes [35] and is the source of detergent resistance of lipid rafts [36]. The latter point is substantiated by the fact that use of MβCD to acutely deplete membrane associated cholesterol results in disruption of lipid rafts [37]. We also found that exposure of N2a murine neuroblastoma cells to MβCD resulted in disruption of lipid rafts and that this disruption was accompanied by an increase in expression of the ganglioside GM1 [13], another lipid associated with rafts. Because lipid rafts are known to function in signal transduction and more specifically are necessary for normal functioning of the oxytocin receptor which can affect behavior [17], it is possible that a generalized disorder

243 in expression of cell membrane associated cholesterol and other raft-associated lipids is associated with onset of developmental disorders identified as autism.

The significant decrease in average cholesterol content of RBC membranes isolated from blood taken from children with autism was accompanied by a similarly significant increase in GM1. Recognizing that CTxB has been reported to bind to a small protein as well as GM1 [38] it is possible that the change in GM1 seen in this study was actually a change in expression of a small peptide. While this possibility needs to be addressed, it would not negate our observation that CTxB binding was on the average enhanced in children diagnosed with autism. Because of its high affinity (found to be 4.61X10-12 using surface plasmon resonance [39]), for GM1, CTxB has been used by many [eg 40] as a cytochemical marker for GM1 in cells known to express it and it has been used in studies of erythrocytes [25, 26]. The presence of GM1 in human erythrocytes, shown chemically by Kundu et al [23], was reported to be the ligand recognized by CTxB when it bound to lipid rafts in erythrocyte membranes [25, 26]. The increase in CTxB binding found using slot blot analysis parallels previous observations indicating that the concentration of GM1, as well as GD1a, GD1b, and GT1b was greater in cerebral spinal fluid from autistic children than in that from controls [19, 41]. An increase in expression of GM1 could account for the reported increase in anti-GM1 autoantibodies found in serum from children diagnosed with autism [20].

The average decrease in cholesterol and increase in GM1 agrees with previous findings that a significant decrease in membrane cholesterol was accompanied by a significant increase in GM1, again identified by binding of CTxB, in neuroblastoma cells [13, 14] and in SH-SY5Y human neuroblastoma cells expressing the H63D variant of the hemochromatosis protein found in lipid rafts [27]. The fact that GM1 is associated with proteins involved in signal transduction (eg.: cAMP kinase [42], and mitogen-activated protein kinase [43]) supports the hypothesis that these changes may be indicative of changes in signal transduction not only in RBCs but in other cells, including those in the central nervous system.

While the results presented support the concept that changes in lipid composition may contribute to the autistic phenotype, they also point out the need to ascertain whether

244 children with autism exhibit changes in expression of proteins involved in cholesterol uptake/release by cells, the expression of enzymes involved in glycosphingolipid metabolism, and whether changes anticipated based on the above results affect expression of proteins known to affect behavior. As more information about the expression and underlying metabolism of cholesterol and glycosphingolipids, major lipid components of membrane associated lipid rafts, by individuals affected with different neurological problems becomes known it should become possible to identify their specific effects. The observations presented point out the need for further investigation of lipid metabolism in autistic subjects the results of which could provide insights into development of new treatment strategies for autism spectrum disorders.

Acknowledgements

This work was supported, in part, under a grant with the Pennsylvania Department of Health’s Health Research Formula Funding Program (State of PA, Act 2001-77-PA Tobacco Settlement Legislation). The Department specifically disclaims responsibility for any analyses, interpretations, or conclusions. Partial financial support for this project was also provided by the Thrasher Research Fund (grant # 02823-0).

245 Figure 6.1: Cholesterol content of RBC membrane. Samples containing 0.1mg of protein were used to extract lipids. A) Values obtained for cholesterol present in RBC membranes isolated from blood from controls compared to those from children with autism, B) cholesterol values versus autism scores. No consistent trend between cholesterol content and autism score was noted. On average samples from children with autism had significantly less cholesterol than those from controls (p<0.05>0.01, unpaired, two-tailed student t-test).

246

Figure 6.2: GM1 expression in the RBC membranes. Relative amount of horseradish peroxidase conjugated CTxB bound to slot blots of RBC membranes isolated from the blood of children with autism relative to those from controls. A) Duplicate blots were prepared using samples of red blood cell membranes containing 2µg of protein isolated from blood obtained from eleven different children with autism and a similar number of samples of blood from controls. GM1 was detected using the HRP-conjugated binding subunit of cholera toxin and bands visualized using Super Signal West Femto Maximum Sensitivity Substrate™ and exposure using a Fuji Film LAS 3000. B) Intensity is given as the (average arbitrary units obtained for duplicates of each sample minus the arbitrary units for the background) X 10-5. Samples from children with autism had significantly more GM1 than those from controls (*, p<0.05>0.01, unpaired, two-tailed student t-test).

247 Figure 6.3: Correlation plot between GM1 (intensity of CTxB) versus cholesterol content of RBC membrane from autistic children. Plot of the intensity of CTxB binding to samples from children diagnosed with autism, squares, and controls, diamonds, versus their cholesterol concentration. Note that while many of the points for children with autism are higher, even though their cholesterol values may be lower, than those for controls.

248 References

1. Mulvihill B, Wingate M, Kirby RS, Pettygrove S, Cunniff C, Meaney FJ, Miller L, Robinson C, et al (2009) Prevalence of autism spectrum disorders–autism and developmental disabilities monitoring network, United States, 2006. MMWR Surveill Summ 58(10):1-20

2. Lord C, Risi S, Lambrecht L, Cook EH Jr, Leventhal BL, DiLavore PC, Pickles A, Rutter M (2000) The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord 30:205-223 doi: 10.1023/A:1005592401947

3. Bu B, Ashwood P, Harvey D, King IB, Van de Water J, Jin LW (2006) Fatty acid composition of red blood cell phospholipids in children with autism. Prostaglandins Leukot Essent Fatty Acids 74:215-221 doi:10.1016/j.plefa.2006.02.001

4. Bell JG, Sargent JR, Tocher DR, Dick JR (2000) Red blood cell fatty acid compositions in a patient with autistic spectrum disorder: a characteristic abnormality in neurodevelopmental disorders? Prostaglandins Leukot Essent Fatty Acids 63:21-25 doi:10.1054/plef.2000.0186

5. Kelley RI (1995) Diagnosis of Smith-Lemli-Opitz syndrome by gas chromatography/mass spectrometry of 7-dehydrocholesterol in plasma, amniotic fluid and cultured skin fibroblasts. Clin Chim Acta 236:45-58 doi:10.1016/0009- 8981(95)06038-4

6. Witsch-Baumgartner M, Fitzky BU, Ogorelkova M, Kraft HG, Moebius FF, Glossmann H, Seedorf U, Gillessen- Kaesbach G, Hoffmann GF, Clayton P, Kelley RI, Utermann G (2000) Mutational spectrum in the ∆7-sterol reductase gene and genotype-phenotype correlation in 84 patients with Smith-Lemli-Opitz syndrome. Am J Hum Genet 66:402–412

249 7. Sikora DM, Pettit-Kekel K, Penfield J, et al (2006) The near universal presence of autism spectrum disorders in children with Smith-Lemli-Opitz syndrome. Am J Med Genet A 140:1511-1518 doi: 10.1002/ajmg.a.31294

8. Tierney E, Bukelis I, Thompson RE, Ahmed K, Aneja A, Kratz L, Kelley RI (2006) Abnormalities of cholesterol metabolism in autism spectrum disorders. Am J Med Genet B Neuropsychiatr Genet 141B:666-668 doi:10.1002/ajmg.b. 30368

9. Saher G, Brugger B, Lappe-Slefke C, Möbius W, Tozawa R, Wehr MC, Wieland F, Ishibashi S, Nave KA (2005) High cholesterol level is essential for myelin membrane growth. Nat Neurosci 8:468-475 doi:10.1038/nn1426

10. Camargo N, Smit AB, Verheijen MH (2009) SREBPs:SREBP function in glia-neuron interactions. FEBS J 276:628-636 doi: 10.1111/j.1742-4658.2008.06808.x

11. Neufeld EB, Cooney AM, Pitha J, Dawidowicz EA, Dwyer NK, Pentchev PG, Blanchette-Mackie EJ (1996) Intracellular trafficking of cholesterol monitored with a cyclodextrin. J Biol Chem 271:21604-21613 doi: 10.1074/jbc.271.35.21604

12. Kitajima Y, Sekiya T, Nozawa Y (1976) Freeze-fracture ultrastructural alterations induced by filipin, pimaricin, nystatin and amphotericin B in the plasma membranes of Epidermophyton, Saccharomyces and red complex-induced membrane lesions. Biochim Biophys Acta 455: 452-465

13. Petro KA, Schengrund CL (2009) Membrane raft disruption promotes axonogenesis in N2a neuroblastoma cells. Neurochem Res 34:29-37 doi: 10.1007/s11064-008- 9625-9

14. Petro KA, Dyer MA, Yowler BC, Schengrund CL (2006) Disruption of lipid rafts enhances activity of botulinum neurotoxin serotype A. Toxicon 48:1035-1045 doi:10.1016/j.toxicon.2006.08.011

15. Petro KA, Schengrund, CL (2010) Lipid rafts: inhibitors of endocytosis of botulinum toxin. The Botulinum J 1:358-374

250 16. Reversi A, Rimoldi V, Brambillasca S, Chini B (2006) Effects of cholesterol manipulation on the signaling of the human oxytocin receptor. Am J Physiol Regul Integr Comp Physiol 291:R861-869 doi: 10. 1152/ ajpregu. 00333. 2006

17. Nishimori K, Takayanagi Y, Yoshida M, Kasahara Y, Young LJ, Kawamata M (2008) New aspects of ocytocin receptor function revealed by knockout mice: sociosexual behaviour and control of energy balance. Prog Brain Res 170:79-90 doi:10.1016/S0079-6123(08)00408-1

18. Gross C, Zhuang X, Stark K, Ramboz S, Oosting R, Kirby L, Santarelli L, Beck S,

Hen R (2002) Serotonin1A receptor acts during development to establish normal anxiety-like behaviour in the adult. Nature 416:396-400 doi:10.1038/416396a

19. Nordin V, Lekman A, Johansson M, Fredman P, Gillberg C (1998) Gangliosides in cerebral spinal fluid in children with autism spectrum disorders. Dev Med Child Neurol 40:587-594

20. Mostafa GA, Al-Avadhi LY (2011) Increased serum levels of anti-ganglioside M1 auto-antibodies in autistic children: relation to disease severity. J Neuroinflammation 8:39 doi: 10.1186/1742-2094-8-39 21. Thomas S, Kumar RS, Casares S, Brumeanu T-D (2003) Sensitive detection of GM1 lipid rafts and TCR partitioning in the T cell membrane. J Immunol Methods 275:161-168 doi:10.1016/S0022-1759(03)00014-0

22. Pang H, Lee PU, Nabi IR (2004) Ganglioside GM1 levels are a determinant of the extent of caveolae/raft-dependent endocytosis of cholera toxin to the Golgi apparatus. J Cell Sci 117:1421-1430 doi: 10.1242/ jcs.01009

23. Kundu SK, Samuelsson BE, Pascher I, Marcus DM (1983) New gangliosides from human erythrocytes. J Biol Chem 258:13857-13866

24. Samuel BU, Mohandas N, Harrison T, McManus H, Rosse W, Reid M, Haldar K (2001) The role of cholesterol and glycosylphosphatidylinositol-anchored proteins of erythrocyte rafts in regulating raft protein content and malarial infection. J Biol Chem

251 276:29139-29329 doi:10.1074/jbc.M101268200 25. Mrowczynska L, Hagerstrand H (2008) Patching of ganglioside (M1) in human erythrocytes – distribution of CD47 and CD59 in patched and curved membranes. Mol Membr Biol 25:258-265 doi:10.1080/09687680802043638 26. Mrowczynska L, Salzer U, Perutkova S, Iglic S, Hagerstrand H (2010) Echinophilic proteins stomatin, sorcin, and synexin locate outside gangliosideM1 (GM1) patches in the erythrocyte membrane. Biochem Biophys Res Commun 401:396-400 doi:10.1016/j.bbrc.2010.09.065 27. Ali-Rahmani F, Hengst JA, Connor JR, Schengrund CL (2011) Effect of HFE variants on sphingolipid expression by SH-SY5Y human neuroblastoma cells. Neurochem Res 36:1687-1696 doi: 10.1007/s11064-011-0403-8 28. Herbert MR (2010) Contributions of the environment and environmentally vulnerable physiology to autism spectrum disorders. Curr Opin Neurol 23:103-110 doi: 10.1097/WCO.0b013e328336a01f 29. Donati RJ, Dwivedi Y, Roberts RC, Conley RR, Pandey GN, Rasenick MM (2008)

lipid raft domains where it is less likely to activate adenylyl cyclase. J Neurosci 28:3042-3050 doi: 10.1523/ JNEUROSCI.5713-07.2008 30. Kamata K, Manno S, Ozaki M, Takakuwa Y (2008) Functional evidence for presence

transduction. Am J Hematol 83:371-375 doi: 10.1002/ajh.21126 31. Corbett BA, Kantor AB, Schulman H, Walker WL, Lit L, Ashwood P, Rocke DM, Sharp FR (2007) A proteomic study of serum from children with autism showing differential expression of apolipoproteins and complement proteins. Molec Psych 12:292-306 doi:10.1038/sj.mp.4001943 32. Kim EK, Neggers YH, Shin CS, Kim E, Kim EM (2010) Alterations in lipid profile of autistic boys: a case control study. Nutr Res 30:255-260 doi:10.1016/j.nutres.2010.04.002 33. Tyler CV, Schramm SC, Karafa M, Tang AS, Jain AK (2011) Chronic disease risks in young adults with autism spectrum disorder: forewarned is forearmed. Am J Intellect Dev Disabil 116:371-380 doi:10.1352/1944-7558-116.5.371

252 34. Aneja A, Tierney E (2008) Autism: the role of cholesterol in treatment. Int Rev Psychiatry 20:165-170 doi:10.1080/09540260801889062 35. Bloch K (1991) Cholesterol: evolution of structure and function. In: Vance DE, Vance JE, (eds) Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier, Amsterdam, The Netherlands. Pp 363-381 36. Brown DA, London E (2000) Structure and function of sphingolipid- and cholesterol- rich membrane rafts. J Biol Chem 275:17221-17224 doi: 10.1074/jbc.R000005200 37. Kabouridis PS, Janzen J, Magee AL, Ley SC (2000) Cholesterol depletion disrupts lipid rafts and modulates the activity of multiple signaling pathways in T lymphocytes. Eur J Immunol 30:954–963 doi: 10.1002/1521- 4141(200003)30:3<954::AID-IMMU954>3.0.CO;2-Y 38. Yanagisawa M, Ariga T, Yu RK (2006) Cholera toxin B subunit binding does not correlate with GM1 expression: a study using mouse embryonic neural precursor cells. Glycobiol 16:19G-22G doi:10.1093/glycob/cwl003 39. Kuziemko GM, Stroh M, Stevens RC (1996) Cholera toxin binding affinity and specificity for gangliosides determined by surface plasmon resonance. Biochem 35:6375-6384 doi: 10.1021/bi952314i 40. Kozireski-Chuback D, Wu G, Ledeen RW (1999) Axonogenesis in neuro-2a cells correlates with GM1 upregulation in the nuclear and plasma membranes. J Neurosci Res 57:541-550 doi: 10.1002/(SICI)1097-4547(19990815)57:4<541::AID- JNR14>3.0.CO;2-X 41. Lekman A, Skjeldal, Sponheim E, Svennerholm L (2008) Gangliosides in children with autism. Acta Pediatrica 84:7877-790 doi: 10.1111/j.1651-2227.1995.tb13757.x 42. Yates AJ, Walters JD, Wood CL, Johnson JD (1989) Ganglioside modulation of cyclic AMP-dependent kinase and cyclic nucleotide phosphodiesterase in vitro. J Neurochem 53:162–167 43. Choi JS, Kim JA, Joo CK (2003) Activation of MAPK and CREB by GM1 induces survival of RGCs in the retina with axotomized nerve. Invest Ophthalmol Vis Sci 44:1747–1752 doi: 10.1167/iovs.01-0886.

253 VITA FATIMA G. ALI-RAHMANI EDUCATION: 2006: Bachelor of Science, (Summa Cum Laude) Biology & Genetics, GPA: 3.98 The University Of Georgia, Athens, GA 2007-2012 Graduate Student in the Molecular Medicine graduate program The Pennsylvania State University, Hershey, PA HONORS & AWARDS 2004-06: Presidential Scholar at UGA 2010: Young investigator Travel Award from American Society for Neurochemistry (ASN) 2011: Science as Art research photo contest winner from ASN 2011: Class of 1971 Endowed Scholarship Award recipient at Penn State College of Medicine 2012: Alzheimer Disease Drug Discovery Foundation (ADDF) Young Investigator Scholar 2012: Dean’s Travel Award at Penn State College of Medicine 2012: 2nd place award for poster presentation-Graduate Student division at PSIN 2012: Class of 1971 Endowed Scholarship Award recipient at Penn State College of Medicine 2012: Graduate Alumni Endowed Scholarship Award recipient at Penn State College PROFESSIONAL & ACADEMIC MEMEBERSHIPS / SOCIETIES: ● Society for Neuroscience (SFN) ● New York Academy of Science ● American Society for Neurochemistry ● American Association for Cancer Research PUBLICATIONS: - Ali-Rahmani, F, Hengst JA, Connor JR, Schengrund CL. Effect of HFE variants on sphingolipid expression by SHSY5Y human neuroblastoma cells. Neurochemical Research. Epub Jan 18th 2011 - Schengrund CL, Ali-Rahmani, F, Ramer JC. Cholesterol, GM1, and Autism. Neurochemical Research. Epub Jan 18th 2012 -F. Ali-Rahmani, P. Grigson, S. Lee, E. Neely, J. Connor, and C-L. Schengrund. Effect of H63D- HFE on cholesterol metabolism, brain atrophy, and cognitive impairment: implications for Alzheimer Disease. Journal of Experimental Medicine (Submitted for review) - F. Ali-Rahmani, M. Huang, S. Lee, C-L. Schengrund, and J. Connor. C282Y-HFE increases cellular proliferation by inducing cholesterol accumulation. (In preparation) - F. Ali-Rahmani, C-L. Schengrund, and J. Connor. Dysregulation of iron and cholesterol metabolism in diseases. (Review; in preparation) Poster Presentations: - Fatima Ali-Rahmani, Sang Lee, James Connor, and Cara-Lynne Schengrund. HFE polymorphism affects cholesterol metabolism: Insights into neurodegenerative diseases. Society for Neuroscience annual meeting 2009- Chicago, IL - F. Ali-Rahmani, S. Lee, J. Connor, and C-L. Schengrund. H63D variant of HFE alters expression of lipid raft components: implications for Alzheimer Disease. American Society for Neurochemistry annual meeting 2010- Santa Fe, New Mexico - F. Ali-Rahmani, S. Lee, W. Nandar, B. Neely, J. Connor, and C-L. Schengrund. Increased expression of cholesterol 24S hydroxylase due to the H63D variant of HFE: a risk factor for Alzheimer Disease. Society for Neuroscience annual meeting 2010- San Diego, CA - F. Ali-Rahmani, S. Lee, W. Nandar, B. Neely, Z. Simmons, J. Connor, and C-L. Schengrund. H63D variant of HFE alters cholesterol metabolism: a potential mechanism for disruption of cell signaling in amyotrophic lateral sclerosis. Motor Neuron Disease-ALS annual meeting 2010- Orlando, FL - F. Ali-Rahmani, S. Lee, J. Connor, and C-L. Schengrund. Analysis of sphingolipid expression: Do HFE variants impact sphingolipid metabolism? American Society for Neurochemistry 2011- Saint Louise, MI

254