INVESTIGATION OF mRNA OXIDATION IN ALZHEIMER’S DISEASE

DISSERTATION

Presented in partial fulfillment of the requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Xiu Shan, Bachelor in Medicine

***********

The Ohio State University

2005

Dissertation Committee: Approved by

Dr. Chien-liang Glenn Lin, Advisor

Dr. Jeff Kuret ------

Dr. James Van Brocklyn Advisor

Dr. Mike Zhu Neuroscience Graduate Studies Program

ABSTRACT

Oxidative modifications to vital cellular components have been described in

association with Alzheimer’s disease (AD). Cytoplasmic RNA oxidation is a prominent feature of vulnerable in AD affected regions. However, the role of

RNA oxidation in the pathogenesis of AD is unknown. This dissertation examined the magnitude of oxidized mRNAs in AD, identified oxidized mRNA species, and characterized the consequence of mRNA oxidation.

An immunoprecipitation method was developed to isolate, characterize and quantify oxidized mRNAs in AD brain. The data showed that 30-70% of mRNAs are oxidized in frontal cortices of AD as compared to age-matched normal control.

Identification of oxidized mRNA species revealed that mRNA oxidation in AD is not

random but selective. Many identified oxidized mRNA species have been implicated

in the pathogenesis of AD. Quantitative analysis revealed that some mRNA species

are more susceptible to oxidative damage for which the relative amounts of oxidized

transcripts reach 50-70%.

To understand whether oxidation of mRNAs is involved in the pathogenesis of

AD, the consequence of mRNA oxidation was studied using in vitro system and cell lines. mRNA oxidation may have deleterious effects by interfering

ii

with normal translational process or by a toxic gain-of-function. Examination of the protein expression from oxidized mRNAs revealed that mRNA oxidation causes a decrease of protein level and associated activity, which may result from the abnormal association of polyribosomes with oxidized mRNAs during the process of translation.

Furthermore, we observed that microinjection of oxidized mRNAs into neuronal cells caused cell death, suggesting that increased mRNA oxidation could be lethal to cells.

Studies in primary neuronal cultures revealed that RNA oxidation is an early event far preceding cell death induced by insults associated with AD. Neurons are specifically susceptible to RNA oxidation. Some mRNA species highly oxidized in

AD brain are also present in oxidized mRNA pool of oxidative insult-treated cultures.

A decrease of protein level was observed to oxidized mRNA species in cultures.

Taken together, these studies indicate the possibility of oxidized mRNAs to interfere with physiological functions of cells and also implicate the potential contribution of RNA oxidation in the pathogenesis of AD.

iii

Dedicated to my parents and sisters

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Chien-liang Glenn Lin for

providing me with an opportunity to work in and learn from his research group. I will

always appreciate his supervision on my graduate study and research. I would also

like to thank the members of my dissertation committee – Dr. Jeff Kuret, Dr. James

Van Brocklyn, and Dr. Mike Xi Zhu – for providing me with many helpful inputs

during the course of my thesis research.

The following individuals are acknowledged for providing me access to their equipments or their inputs related to said equipments: Dr. Georgia A. Bishop

(Department of Neuroscience) and Dr. James S. King (Department of Neuroscience) for access to the Zeiss inverted microscope, Dr. David Suffen (Department of

Pharmacology) and Ju Young Kim (Department of Pharmacology) for access to the

SW60 ultracentrifuge rotor, Dr. Richard W. Burry (Department of Neuroscience) and

Kathy Wolken (Campus Microscope and Imaging Facility) and Brian Kemmenoe

(Campus Microscope and Imaging Facility) for access to the Zeiss laser-scanning confocal microscope.

v

I thank Johns Hopkins Alzheimer’s Disease Research Center Brain Bank and ALS

Brain Bank, Harvard Brain Tissue Resource Center and University of Maryland Brain and Tissue Banks for Developmental Disorders for providing AD and normal control brain tissues.

I would like to thank the current and past members of the Lin research group, Dr.

Hirofumi Tashiro, Dr. Matthew E. R. Butchbach, Dr. Hong Guo, Ms. Liching Lai, Ms.

Yueming Chang, Ms. Guilian Tian and Mr. Michael P. Stockinger, for their support during the course of my thesis research. Especially, I would like to thank Dr.

Hirofumi Tashiro for sharing the project with me and Dr. Matthew E. R. Butchbach

for providing me many helpful insights for my experiment and vigorous discussion for my manuscripts and kind suggestion for my postdoctoral position searching.

I would like to thank our collaborator, Dr. Chandan Sen and Dr. Soma Chaki for helping us with the cell microinjection experiment.

Finally, I would like to thank my family for their constant support throughout my education

vi

VITA

August 24, 1977…………………..Born – Hebei Province, P.R.China Sept 1995 - June 2000 Bachelor in Medicine Beijing Medical University, Beijing, P.R.China May 1999 - June 2000 Undergraduate Research Assistant, Neuroscience Research Institute, Beijing Medical University, P.R.China Sept 2000 – present Graduate Research Associate Neuroscience Graduate Studies Program The Ohio State University

PUBLICATIONS

1. Shan X, Tashiro H, and Lin CG (2003) The Identification and Characterization of Oxidized RNA in Alzheimer’s Disease (J Neurosci. 23:4913-4921)

2. Guo H, Lai L, Butchbach MER, Stockinger MP, Shan X, Bishop GA and Lin CG (2003) Increased expression of the glial glutamate transporter EAAT2 modulates excitotoxicity and delays the onset but not the outcome of ALS in mice. (Human Mol Genet 19:2519-2532)

FIELDS OF STUDY

Major field: Neuroscience

vii

TABLE OF CONTENTS

Page

ABSTRACT…...…………………………..………………………………………. ii

ACKNOWLEDGMENTS…..…………………………...……………………….. v

VITA……………………...………………...…………………..…………………. vii

LIST OF TABLES…………………..…………….………………………..…….. xi

LIST OF FIGURES………………………..…………………………………….. xii

LIST OF SYMBOLS AND ABBREVATIONS……………………..…………...xiv

Chapter 1 INTRODUCTION…………………………………………………… 1 1.1 Alzheimer’s disease…………………………………………..……………. 1 1.2 regulatory pathway………………………..…….. 7 1.3 hypothesis in Alzheimer’s disease………..……………… 9 1.4 Markers for oxidative stress………………………………………….…….. 18 1.5 Lipid peroxidation in Alzheimer’s disease………………………………….19 1.6 Protein oxidation in Alzheimer’s disease…………………………..…..….. 22 1.7 DNA oxidation in Alzheimer’s disease………………………………..……25 1.8 RNA oxidation in Alzheimer’s disease…………………………..….……...29 1.9 Oxidative stress in other neurodegenerative diseases………………..…….. 30

Chapter 2 QUANTIFICATION OF OXIDIZED mRNAS IN ALZHEIMER’S DISEASE..…………………..…………………………………………………….. 36

2.1 Abstract…………………..……………………...………………..……….. 36 2.2 Introduction………………..……………………………………………… 37 2.3 Materials and Methods…………………..………………………………..38 2.4 Results…………………….…………………………………………...…... 45 2.4.1 The quality of RNA from postmortem tissues………………….……… 45 2.4.2 Detecting of oxidized mRNA in AD by Northern blotting………….….46 2.4.3 Immunoprecipitation of oxidized mRNAs….…………………………..47 2.4.4 Quantification of the level of oxidized mRNAs in diseased and normal tissues……………………………..………………………………….……..... 52 2.5 Discussion………………………..………………...……………………… 54

viii

Chapter 3 CHARACTERIZATION OF OXIDIZED mRNA SPECIES IN ALZHEIMER’S DISEASE………………………………..……………………... 58

3.1 Abstract...... 58 3.2 Introduction……………………………………..………………………… 58 3.3 Materials and Method………………………………………..……………60 3.4 Results……………………………………………………………………... 64 3.4.1 and sequence analysis of oxidized mRNAs from Alzheimer’s brains……………………………………..………...……………64 3.4.2 Filter array to examine the susceptibility of individual mRNA species to oxidation in AD.……………………….……………………………………...70 3.4.3 Quantification of the level of oxidation for individual mRNA species by filter array and semi-quantitative RT-PCR………………………….……….. 72 3.4.4 Characterization of the selectivity of oxidation to individual mRNA species………………………………………………………………………... 75 3.5 Discussion…………………..…………………………………………...… 75

Chapter 4 THE CONSEQUENCE OF mRNA OXIDATION…………..…….. 81

4.1 Abtract………………………………………………………..…………… 81 4.2 Introduction…………………………………………………..…………… 82 4.3 Materials and Methods………………………………………………..….. 84 4.4 Results...... 91 4.4.1 RNA oxidation causes decreases of expression level and activity of the protein in rabbit reticulocyte lysate………………………………..…………. 91 4.4.2 RNA oxidation causes decreases of expression level and activity of the protein in cell lines………………………………………..………………….. 93 4.4.3 RNA oxidation causes the abnormal increase of polyribosome association……………………………………………………………………. 98 4.4.4 No truncated protein was generated from oxidized mRNA examined by immunoprecipitaion and Western blotting ……………..…………………… 102 4.4.5 Investigation of protein from oxidized mRNA by SELDI- TOF…………………………………..………………………………… 104 4.4.6 mRNA oxidation could be lethal to cells……………………..………... 105 4.5 Discussion………………………………………..………………………... 108

Chapter 5 INVESTIGATION OF RNA OXIDATION IN CELL MODEL….. 115

5.1 Abstract…………………………………………….……………………....115 5.2 Introduction………………………………………….……………………. 116 5.3 Methods…………………………………………………………..………... 118 5.4 Results…………………………………………………………………..…. 123 5.4.1 The toxic effects of different insults associated with AD to neurons..… 123

ix

5.4.2 H2O2 treatment induces RNA oxidation in the early stage of neuronal cell death…………………………………………..……………………………… 126 5.4.3 RNA oxidation is increased in neuronal cultures treated with glutamate and Aβ25-35……………………………………………...………………….. 129 5.4.4 RNA oxidation is primarily observed in neurons, not in astrocytes..….. 131 5.4.5 Identification of oxidized mRNA species and further investigation of possible consequence of oxidized mRNA species in H2O2-treated neuronal cultures……………………………………………………………………….. 132 5.5 Discussion...... 134

Chapter 6 CONCLUSIONS AND PERSPECTIVES………………..…………. 140

List of references………………………………………………………………….. 144

x

LIST OF TABLES

Table 1. Classification of Clones and statistical analysis on filter array…………….66 Table 2. The magnitude of oxidation of specific mRNAs in AD frontal cortices…...74

xi

LIST OF FIGURES

Figure 1.1 Flow chart to show the oxidative stress hypothesis in the pathogenesis of

Alzheimer’s disease………………………………………………………………... 10

Figure 1.2. Immunohistochemical method to detect oxidized in AD.... 29

Figure 2.1 Northwestern analysis to visualize oxidized mRNA…………………… 47

Figure 2.2 Immunoprecipitation protocol to isolate oxidized mRNA……………... 48

Figure 2.3 Southern blotting analysis of 15A3-immunoprecipitated oxidized mRNAs…………………………………………………………………………….. 51

Figure.2.4 Significant amounts of mRNAs are oxidized in AD frontal cortices…... 53

Figure 3.1 Electrophoresis of PCR-amplified cDNA products……………...…….. 65

Figure 3.2 PCR analysis of identified clones………………………………………..69

Figure 3.3 Filter array analyses of identified oxidized mRNA species reveal that some mRNA species are more susceptible to oxidative damage…………………………..71

Figure 3.4 Quantification of the magnitude of oxidation for individual mRNA species by filter array and semi-quantitative RT-PCR……………………………………….73

Figure 4.1 Oxidation of luciferase mRNAs cause s the decrease of functional activity and expression level of the protein……...………………………………… .93

Figure 4.2 Expression of oxidized luciferase RNA in HEK cells…………………. .94

Figure 4.3 Expression of EGFP mRNA in HEK293 and Neuro2A cell lines……….95

Figure 4.4 Oxidation of hSOD1 mRNA results in a decrease of detectable protein level of hSOD1 in HEK293 cells…………………………………………………....96

xii

Figure 4.5 Detection of stability of oxidized and non-oxidized mRNAs at different

time points after transfection by quantitative RT-PCR analysis…………………… 98

Figure 4.6 Polyribosome profiles of non-oxidized and oxidized luciferase mRNAs in

rabbit reticulocyte lysates………………………………………………………….. 101

Figure 4.7 Western blotting to examine the presence of truncated proteins translated

from oxidized luciferase mRNA………………………………………………..….. 103

Figure 4.8 Delivery of oxidized EGFP mRNAs into HT4 cells causes cell death… 106

Figure 4.9 Delivery of oxidized mRNAs of HT4 cells causes cell death…...……... 108

Figure 5.1 MAP2 immunostaining to examine -toxic effects of different insults

in primary neuronal cultures…………………………………………………..…… 125

Figure 5.2 Immunostaining with 15A3 to examine the time course of RNA oxidation

in H2O2-treated neuronal cultures……………………………………………..…… 126

Figure 5.3 Examination of cell viability by nuclear staining and LDH release assay………………………………………………………………………..………. 128

Figure 5.4 Immunoprecipitation of oxidized mRNAs in H2O2-treated neuronal

cultures………………………………..………………………………….………… 129

Figure 5.5 RNA oxidation in glutamate- or Aβ-treated neuronal cultures………… 130

Figure 5.6 Localization of oxidized RNAs in neuronal cells………………………. 131

Figure 5.7 Identification of oxidized mRNA species in H2O2-treated neuronal cultures by RT-PCR………………………………………………………………………… 133

Figure 5.8 Examination of the protein and mRNA levels of oxidized mRNA species in neuronal cultures………………………………………………………………… 134

xiii

LIST OF SYMBOLS AND ABBREVATIONS

Aβ amyloid β peptide AD Alzheimer’s disease AGEs advanced glycation end products ALS amyotrphic lateral sclerosis APOE apolipoprotein E APP amyloid precursor protein ATP adenosine triphisphate BACE β-site APP cleaving BSA bovine serus albumin Cbm cerebellum cDNA complementary DNA CHCA cyano-4-hydroxy cinnamic acid CIDE-B cell death-inducing DFF45-like effector B COX cytochrome c oxidase CSF cerebrospinal fluid Cu copper DEPC diethylpyrocarbonate DHA docosahexaenoic acid DIG digoxogenin DMEM Dulbecco’s modified Eagle medium DTT dithiothreitol EAAT3 excitatory amino acid transporter 3 EGFP enhanced green fluorescence protein FAD familiar AD FBS fetal bovine serum FC frontal cortex

xiv

Fe iron GDH glutamate dehydrogenase 1 GFAP glial fibrillary acid protein GSH reduced glutathione GSH-Px glutathione peroxidase GSSG oxidized glutathione GSSG-R glutathione reductase HD Huntngton’s disease HEK human embryonic kindey Hip HNE 4-hydroxynonenal

H2O2 hydrogen peroxide hr hour(s) HRP horseradish peroxidase hSOD1 Cu/Zn superoxide dismutase IgG immunoglobulin G iNOS inducible nitric oxide synthase IsoFs isofurans IsoPs isoprostanes kD kilodalton LDH lactate dehydrogenase mAb monoclonal antibody MAO monoamine oxidase MAP2 microtubule associated protein 2 MCI mild cognitive impairment MDA malondialdehyde min minute(s) MPTP 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine

xv

mRNA messenger RNA MS-MS Tandem Mass Spectrometry mtDNA mitochondrial DNA NFT neurofibrillary tangles nNOS neuronal nitric oxide synthase NO nitric oxide

O2˙¯ superoxide OH˙ 8-OHdG 8-hydroxy-2'- 8OHG 8-oxo-7,8-dihydroguanosine ONOO˙) peroxynitrite pAb polyclonal antibody PAGE polyacrylamide PBS phosphate buffer saline PCR polymerase chain reaction PD Parkinson’s disease PHF paired helical filaments PS presenilin PVDF polyvinylidene difluoride RNS reactive nitrogen species ROS reactive oxygen species RT-PCR reverse transcriptase-polymerase chain reaction SDS sodium dodecyl sulfate SELDI-TOF surface-enhanced laser desorption and ionization-time of flight mass spectrometry SOD superoxide dismutase SPA sinapic acid TBARS thiobarbituric acid-reactive substances

xvi

TNF- tumor necrosis factor- UTR untranlational region vCSF ventricular cerebrospinal fluid YB-1 Y-box-binding protein 1

xvii

CHAPTER 1

INTRODUCTION

1.1 Alzheimer’s disease

1.1.1 Introductary remarks

Alzheimer’s disease (AD) is the most common cause of adult dementia and is

the fourth leading cause of death in developed nations (after heart disease, cancer, and

stroke). Alzheimer’s patients show a gradual loss of and cognitive functions.

As the disease progresses, patients may experience changes in personality and behaviors. AD develops at various rates. The course duration of the disease varies from three to twenty years (McKhann et al., 1984).

Aging is the major contributor to the onset of the disease. Although some genetic and other risk factors have been discovered to be associated with the disease, a large portion is sporadic. Because of the aging population, AD has become more common and has been a major health problem in the United States. It is estimated that up to 4 million people currently develop the disease. One out of every 10 people 65 years and older is afflicted with AD, and about half of Americans 85 years and older suffer

1

from AD. The prevalence of the disease doubles every 5 years beyond age 65. By

2050, the number could increase to 45 million.

No effective treatments have been discovered. The costs associated with the disease, including health care expenses and lost wages of both patients and their

caregivers, load a heavy economic burden on families and society. This makes AD an

urgent priority in research. Recent studies on AD are focusing on discovering the

pathogenesis of the disease, exploring better ways to diagnose the disease in the early

stage, and developing drugs that may help slow down the progress of the disease and

treat symptoms.

1.1.2 Neuropathological changes

Since the first report of AD case in 1907, knowledge of the disease has been

expanding rapidly, especially in past two decades. The hallmark pathohistological

changes in AD include amyloid plaques, neurofibrillary tangles (NFT), and a large

scale loss of synapse and neurons. The diagnosis of AD can only be made on autopsy

by the presence of those changes in specific brain areas, including hippocampus,

entorhinal cortex, neocortex, nucleus basalis, and several brain stem nuclei (Arnold et al., 1991). Amyloid plaques are extracellular protein aggregates with other complex

structures containing dystrophic neurites, reactive astrocytes and microglia. Amyloid

β peptide (Aβ), a 4kD, 39-43 amino acid peptide derived from the cleavage of

amyloid precursor protein (APP), is the major component of plaques. Amyloid

plaques typically begin first in the association areas of cerebral cortex. Compared to

2

NFT, amyloid plaque correlates less closely with the severity of the disease. NFTs are

inclusions within cell bodies, proximal dendrites, and filamentous swelling in distal axons. Tangles are formed by hyperphosphorylated microtubule-associated protein tau, which assemble into paired helical filaments (PHF). NFTs typically appear first in the entorhinal cortex. On a cellular level, pyramidal neuron involved in corticocortical pathways is the major cell type lost in AD brain.

1.1.3 Pathogenesis of AD

Amyloid cascade hypothesis. The pathogenesis of AD is still under investigation.

Some hypotheses on the cause of the disease are proposed based on the pathological changes in Alzheimer’s brains. Amyloid cascade hypothesis states that the abnormal processing of APP triggers an increased production of Aβ and a series of downstream events leading to neurodegeneration. APP is a type I transmembrane protein expressed highly but not exclusively in neurons. The normal function of APP is not well-known. APP is proposed to have a neuroprotective effect against neuronal stress or injury by reducing Ca2+ concentration and protecting neurons from glutamate

excitotoxicity (Mattson et al., 1993). Injections of a 17-peptide subunit of APP had

been shown to reduce ischemic damage on neurons significantly (Bowes et al., 1994).

α-, β-, and γ-secretases are three proteolytic to process APP. Pathogenic Aβ

peptides, Aβ40 and Aβ42, especially the latter one, are generated from cleavage of

APP sequentially by β-secretase and then γ-secretase. Cleavage of APP within Aβ

domain by α-secretase precludes the Aβ formation. The protein BACE (β-site APP

3

cleaving enzyme) has been identified as β-secretase in 1999 (Vassar et al., 1999). γ- secretase has recently been fully characterized as a protein complex composed of presenilin, nicastrin, and APH-1 and PEN-2. Typically α-secretase acts on the cell

surface, whereas β-secretase cleaves at the endoplasmic reticulum. γ-secretase generates Aβ42 if cleavage occurs in the endoplasmic reticulum and Aβ40 if the cleavage is in the trans-Golgi network (Hartmann et al., 1997). Aß42, compared to

Aβ40, is more toxic to neurons due to its higher tendency to form insoluble Aβ

aggregates. Extracellular accumulation of Aβ, especially Aβ42, promotes the

formation of aggregates, fibrils and eventually amyloid plaques. It has been proved

that the multimer and fibril forms of Aβ are toxic to neurons via the following

mechanisms. (1) Aβ can form a cation-selective ion channel after incorporating into cell membrane; (2) Aβ can regulate calcium influx through voltage-gated calcium channels and NMDA receptors; (3) Aβ can regulate the production of reactive oxygen species (ROS).

This amyloid cascade hypothesis is supported in part by the evidence from the studies of familial AD (FAD). Approximately 10% of AD cases are familial cases inherited in an autosomal dominant manner. In FAD, the onset of the disease is between age 40-50s, therefore it is called early-onset AD. To date, in three have been found to occur in FAD. They are APP on chromosome 21,

Presenilin 1 (PS1) on chromosome 14, and Presenilin 2 (PS2) on chromosome 1

(Price et al., 1998). PS1 and PS2 are cofactors of the γ-secretase complex. Many APP mutations found in FAD are close to the cleavage sites involved in the formation of

4

Aβ. The Swedish mutation (K670N, M671L), located at the cleavage site of BACE1,

facilitates the activity of BACE1 and promotes the production of Aβ40. The APP 717

mutation (V717F, V717G, or V717I) is close to the cleavage site of γ-secretase and

facilitates its activity leading to the increased secretion of Aβ42. The PS genes have

been reported to harbor about 100 different FAD mutations. Mutations in PS proteins generally affect the activity of γ-secretase and therefore cause a increase of Aβ42

production.

Tau hypothesis. Transgenic mice with mutant human APP and/or PS do not develop

full neuropathological changes resembling Alzheimer’s brain. NFTs are not present in

any of these transgenic mouse models. Since the number of neocortical NFTs

correlates more closely with the dementia severity in AD patients than that of

amyloid plaque, some researchers believe that NFT of tau protein may play an

important role in the pathogenesis of AD. Tau is a member of the family of

microtubule-associated proteins. There are six isoforms of tau protein differentially expressed in neurons. The normal function of tau protein is to stabilize axonal microtubules and regulate intracellular vesicle transport. Hyperphosphorylation of tau associated with AD impairs its ability to regulate microtubule assembly and promotes the protein aggregation into PHF. However, the encoding tau protein is located on chromosome 17 and is not associated with any FAD. Mutations in tau cause the development of autosomal dominant tauopathy “frontotemporal dementia and

parkinsonism linked to chromosome 17” instead of AD (Hutton et al., 1998). The

5

abnormal hyperphosphorylation of tau could be associated with overactivity of some kinases responsible for tau phosphorylation and/or underactivity of that

normally dephosphorylate tau protein.

Other genetic and enviormental risk factors. As mentioned earlier, most cases of

AD are sporadic and have the onset after age 65 (late-onset). Several genetic risk

factors have been discovered to be associated with late-onset AD. The gene encoding

cholesterol-carrying apolipoprotein E (APOE) on chromosome 19 has been linked to

late-onset AD (Corder et al., 1993). APOE has three alleles: APOE E2, E3 and E4.

APOE E4 dose is positively correlated with the increased risk and earlier onset of the

disease. APOE E2 allele has a protective effect. 40% of AD patients have one or two

APOE E4 alleles. Other genes that could be involved in AD include α-macroglobulin,

a gene encoding a component of α-ketoglutarate dehydrogenase complex, K-variant

of butyl- and several mitochondrial genes. Gender and a history of head

injury or cardiovascular disease influence the onset and progression of the disease.

Oxidative stress hypothesis. Oxidative stress is believed to be an important factor in

normal aging and neurodegenerative diseases including AD, Parkinson’s disease (PD)

and amyotrophic lateral sclerosis (ALS). Free radical is any molecular species with an

unpaired electron which is highly reactive with other compounds such as proteins,

lipids and nucleic acids. Those species are normally generated from tricarboxylic acid

cycles in mitochondria during glycolysis. Brain is very susceptible to oxidative stress

6

because of its high consumption of oxygen in metabolic pathway, abundant lipid

content, and relative paucity of antioxidant enzymes compared to other tissues.

Although representing only 2% of body mass, brain uses 20% of glucose and oxygen

supplied to body. Postmitotic cells, such as neurons, accumulate oxidatively damaged

products through the entire life. Neurons may be the main victim from oxidative

stress partially by this reason. In diseased conditions, the production of free radicals is

significantly increased, therefore leading to pathological changes in neuronal cells.

Details in how oxidative stress occurs and is involved in the diseases will be

described in following sections.

Every hypothesis mentioned above is feasible. However none of them is patently

complete for the pathogenesis of AD. They may work together to generate the

heterogeneous nature of AD.

1.2 Regulatory pathways of reactive oxygen species

Free radicals include superoxide radical (O2˙¯), hydroxyl radical (OH˙) and peroxynitrite (ONOO˙). Hydrogen peroxide (H2O2) and nitric oxide (NO) are not

radicals because of the absence of an unpaired electron, but they can form highly

reactive species, hydroxyl radical and peroxynitrite respectively. Together with free radicals, H2O2 and NO belong to reactive oxygen/nitrogen species (ROS/RNS). ROS

are normal metabolic byproducts of cells. They possess physiological functions under

certain conditions. Normally they are detoxified by intracellular antioxidant defense

systems. There is a balance between the formation and removal of intracellular ROS

7

under normal condition so that cells are protected from the damaging effects of these

substances. Mitochondrium is the major site for the generation of free radicals. The

electron transfer from the donors produced in the tricarboxylic acid cycle to

molecular oxygen causes proton translocation and energy release for ATP synthesis.

Electrons can leak out from the energy-transducing sequences and then bind with molecular oxygens to form superoxide radicals in the normal process of mitochondrial respiration. Superoxide radicals have a dual effect. They are involved in cell defense pathway against infectious microorganisms, but they can also damage cells by themselves or forming the most reactive species hydroxyl radicals or peroxynitrite

Superoxide radicals cross cell membrane through anion channels. They can be converted into hydrogen peroxide by superoxide dismutase (SOD) as shown in reaction [1]. There are two types of SOD: one is manganese-dependent (Mn SOD) located in mitochondria, where it directly converses superoxide radical derived from the electron transfer chain into hydrogen peroxide. The other is copper- and zinc- dependent (Cu/Zn SOD) located in the cytosol with a more generic catalytic function.

Hydrogen peroxides can also be produced from the oxidative deamination of monoamines catalyzed by monoamine oxidase (MAO), which is associated with the outer mitochondrial membrane.

SOD + 2 O2˙¯ + 2H H2O2 + O2 [1]

8

Catalase and glutamate peroxidase are two major enzymes in the cytoplasm to

decompose intracellular hydrogen peroxides (see reaction [2]). Glutathione

peroxidase converts reduced glutathione (GSH) and hydrogen peroxide into oxidized

glutathione (GSSG) and water. GSH has an antioxidant characteristic.

2H2O2 2H2O + O2

Glutathione peroxidase 2GSH + H2O2 2H2O + GSSG [2]

Although hydrogen peroxide itself is not reactive, it can diffuse freely within a long distance and form hydroxyl radical in the presence of redox metal ions by the

Fenton reaction (reaction [3]). There is no enzyme to remove intracellular hydroxyl radicals. Therefore, the only protective strategy against hydroxyl radical is to

minimize its formation.

H O + Fe(II) or Cu(I) OH˙ + OH ¯ + Fe(III) or Cu(II) [3] 2 2

1.3 Oxidative stress hypothes is in Alzheimer’s disease

Increasing evidence demonstrates that the development of oxidative stress, in

which the production of ROS overwhelms antioxidant defense, and the resulting

toxicity are involved in neurodegeneraiton in AD. Some pathways are summarized in

Figure 1.

9

Excitotoxicity APP mutation Presenilin mutation Mitochondrial dysfunction

Disrupted Ca2+ Aβ aggregate Homeostasis

Aging Amyloid Head injury Oxidative deposition Ischemia stress

Redox metal imbalance

Lipid peroxidation Protein oxidation DNA/RNA oxidation

Synaptic damage Cytoskeletal alteration Proteasome dysfunction Apoptosis

NEUROTOXICITY

Figure 1.1 Flow chart to show the oxidative stress hypothesis in the pathogenesis of Alzheimer’s disease.

10

Alteration of redox metal ion homeostasis in AD. As previously discussed, redox

metal ions, especially copper (Cu) and iron (Fe) in their reduced forms, play an

essential role in the formation of ROS. Disruption of Cu and Fe metabolism has been

found in AD brain. In 1953, an increase of iron in the cortex of AD was detected by a

histochemical approach (Goodman, 1953). There is a significant elevation of iron in

NFT-bearing neurons compared to NFT-free neurons in AD (Good et al., 1992).

Smith et al. (1997) found that iron is aberrantly accumulated in cytoplasm in

vulnerable neurons in AD brain. Iron-binding proteins including ferritin and

transferrin are also dysregulated in AD. Ferritin, primarily present in microglia as an

iron-storage protein, is greatly increased in these cells around neuritic plaques and

blood vessels (Connor et al., 1992; Grundke-Iqbal et al., 1990). Transferrin, an iron-

transporting protein, has been found significantly elevated in the cerebrospinal fluid

(CSF) and brain of AD (Loeffler et al., 1995). The transferrin C2 variant is associated

with late-onset AD (Namekata et al., 1997; van Rensburg et al., 2000). In the diseased

conditions, iron may be mobilized from its binding proteins and participate in oxidative stress in cells. Copper has been found in abnormally high concentrations in amyloid plaques and affected neuropils in AD brains.

Copper ions may promote aggregation of Aβ and/or stabilization of Aβ fibrils.

Studies showed that APP, or a synthetic peptide representing the Cu- of

APP, can reduce Cu2+ to Cu+ and then facilitate hydroxyl radical production

(Multhaup et al., 1996). Ceruloplasmin, a copper-binding protein that mediates the

entry of copper into brain, is increased in CSF and brain of AD. However, the level of

11

ceruloplasmin in neurons is not changed, which may leave neurons more vulnerable

to the oxidative effect of increased copper.

Mitochondrial abnormalities in AD. Mitochondria are both the source of ROS and

target of oxidative damage. Dysfunction of proteins in mitochondrial electron

transport chain has been associated with the pathophysiology of AD (Blass and

Gibson 1991). It was shown that cytochrome oxidase activity in AD brains was

compromised. Kish et al. (1992) measured a 16-26% decline in the activity of

complex IV in the cerebral cortex of AD subjects as compared to the age-matched

normal controls. Studies on mitochondria isolated from AD brain demonstrated a

decreased activity of the heme-containing components of complex IV but no change

in their protein levels, suggesting that the decreased enzyme activity is due to abnormal catalytic function rather than decreased enzyme level (Parker et al., 1994).

Inhibition of complex IV activity in submitochondrial particles leads to the increased production of superoxide radicals (Partridge et al., 1994). The decrease of complex

IV acitivity could cause diversion of electrons from their normal pathway into reaction with molecular oxygen, therefore resulting in increased generation of superoxide radicals (Markesbery 1997). Moreover, studies with cytoplasmic hybrid cells, in which mitochondria from sporadic AD were fused with some cells, indicated a deficit in mitochondria function in AD (Ghosh et al., 1999). Using the techniques of in situ hybridization, immunohistochemistry, and electron microscopy, Smith and his colleages found an increase of mitochondrial DNA (mtDNA) and protein but a

12

significant decrease of the number of mitochondria in the cytoplasm of vulnerable

neurons in AD brain (Hirai et al., 2001). The increased mtDNA and protein were also

found in lipofuscin, a lysosome that is the site for mitochondrial degradation. As

suggested by the arthors, these findings indicate that vulnerable neurons in AD have a reduction of proteolytic turnover, leading to accumulation of mtDNA and mitochondrial proteins. In cell culture system, PC12 cells bearing the Swedish double mutations in APP gene have a substantial Aβ level but a reduced adenosine

triphisphate (ATP) level. In 3-month-old Swedish mutant hAPP transgenic mice, which exhibit no plaques but have detectable Aβ levels in the brain, reduced ATP levels can also be observed with a significantly decreased mitochondrial membrane potential under basal conditions (Keil et al., 2004). Indeed, a decrease in mitochondrial metabolic performance and ATP production is strongly associated with apoptosis (Green and Reed 1998). As one of the hallmarks in AD, NFT can be regulated by the level of ATP (Blass et al., 1990). An insufficient ATP production can cause the formation of abnormal cytoskeletal products. Impaired mitochondrial

ATP synthesis activates the brain-specific protein kinase PK 40 erk that phosphorylates tau and results in the generation of paired helical filaments (Roder et al., 1993).

Abnormal activities of antioxidant enzymes in AD. The major enzymes in intracellular antioxidant system include SOD, catalase, and glutathione peroxidase

(GSH-Px ) and reductase (GSSG-R). Results from the studies on activities of

13

antioxidant enzymes in AD brains are somewhat controversial. But dysregulation of antioxidant enzymes by either increasing or decreasing their activities can cause the production of ROS. For example, hydrogen peroxide will accumulate if the ratio of

SOD to catalse increases, resulting in the formation of hydroxyl radicals due to excessive substrates. Marcus et al. (1998) showed that SOD activity was significantly decreased in frontal and temporal cortices of AD, and catalase activity was significantly decreased in AD temporal cortex. This supports the theory that AD brains are experiencing oxidative stress, most significantly in temporal cortex where the pathophysiologic changes are most severe. No significant difference in GSH-Px activity was found in the frontal, temporal, or cerebellar cortex of AD. Lovell et al.

(1995) had a different observation from AD brains. They found that a significant elevation of GSH-Px and GSSG-R activities in AD hippocampus compared to control.

Catalase activity was significantly elevated in AD hippocampus and superior and middle temporal gyri. SOD levels were elevated in all brain regions in AD patients.

Authors suggested that it is a compensatory rise of antioxidant activities in response to increased ROS formation in AD.

Apo E and oxidative stress. ApoE E4 is linked to late-onset AD, whereas apoE E2 offers a protective effect. Apo E functions as a cholesterol-transporting protein in both the vascular system and neurons. As a crucial cholesterol carrier, apoE is involved in cholesterol-dependent functions, such as dendritic remodeling and synaptogenesis in which changes to membrane lipids are required. Poirier (1994)

14

proposed that apoE performs a beneficial effect to neurons but that E4 isoform is less

effective than are E2 and E3 isoforms. Ramassamy et al. (1998) established a relation

in AD between apoE genotype and lipoperoxidation, a marker for oxidative damage

on lipids. They showed that the level of peroxidation in AD brain is related to apoE

genotype. E4 isoform associated with a higher level of peroxidation compared to E2

and E3. Also the level of lipoperoxidation in AD is inversely proportional to the

concentration of apoE. This result supports the hypothesis that oxidative stress is

involved in the pathogenesis of AD in association with apoE genotype.

Glial inflammation and oxidative stress in AD. Reactive astrocytes and microglia

are present in the periphery of amyloid plaques in AD brain. Glial cell activation

involves a complex range of responses, including increased expression of genes

responsible for the syntheses of NO and proinflammatory cytokines (Colton and

Gilbert 1987; Klegeris and McGeer 1994). Activated microglia kill neurons in vitro

by a variety of released substances, including NO (Boje and Arora 1992; Meda et al.,

1995), superoxide radicals (Tanaka et al., 1994), and tumor necrosis factor- (TNF- )

(Wood 1995). The interaction of NO with superoxide forms the neurotoxic

peroxynitrite, which can result in nitration of tyrosine and nitrosylation of cysteines

within both enzymes and structural proteins. Widespread peroxynitrite-mediated

damages are present in affected areas of AD brain (Smith et al., 1997; Koppal et al.,

1999)

15

More recently, studies showed that Aβ can elicit its neurotoxic effect by activating microglia to induce the formation of peroxynitrites (Xie et al., 2002).

Peroxynitrite acts through multiple mechanisms including the induction of proton leakage across the inner mitochondrial membrane, inhibition of mitochondrial enzymes, and damage to antioxidant enzymes (Merry 2002).

Role of Aβ in ROS generation. Increasing evidence indicates that the cytotoxicity of Aβ is mediated by free radical generation. Micromolar concentration of Aβ can cause hydrogen peroxide accumulation in neuroblastoma cells in a time- and concentration-dependent manners (Behl et al., 1994). Cells selectively resistant to Aβ are also highly resistant to hydrogen peroxide. Catalase can block Aβ toxicity indicating that hydrogen peroxide is the mediator (Behl et al., 1994). Aβ also potentiates the toxicity of hydrogen peroxide and iron-induced oxidative injury in neuronal cultures (Saunders et al., 1991). Thomas et al. (1996) demonstrated that Aβ causes the damages in vascular endothelial cells which are prevented by the enzyme

SOD. This suggests that Aβ may generate superoxide radicals to mediate toxicity.

Synthetic Aβ can inactivate glutamine synthetase and creatine kinase, which are susceptible to oxidative damage, in aqueous solution and in cell cultures only in the presence of oxygen (Hensley et al., 1994). Antioxidants, such as vitamin E and spin- trap compound (PBN), can protect cells against Aβ toxicity (Harris et al., 1995). All of these studies show that Aβ stimulates the production of ROS directly or indirectly.

Aβ is capable of mediating oxidative stress via several pathways as followings. (1)

16

Aβ can impair the activities of all mitochondrial respiratory chain complexes (Parks et

al., 2001), more severely cytochrome c oxidase (COX) (Parker et al., 1994; Canevari et al., 1999). Incubation with Aβ decreased ATP concentration to 58% of control in neurons and 71% of control in astrocytes (Casley et al., 2002). (2) Aβ can cause

mitochondrial structural damages (Hashimoto et al., 2003) and decrease total

mitochondrial number (Casley et al., 2002). (3) Aβ can impair endogenous

antioxidant systems (Kim et al., 2003). Levels of reduced glutathione, Mn SOD,

glutathione peroxidase and glutathione-S- were lowered by Aβ in neurons

and/or astrocytes. (4) Aβ can inhibit glutamate uptake leading to overstimulation of

NMDA receptors by glutamate, which causes an increase of intracellular Ca2+ level.

Activation of the enzyme secondary to the increase of Ca2+ may be involved in the

formation of ROS (Facheris et al., 2004). (5) Aβ can activiate microglia therefore

upregulate the level of nitric oxide synthase leading to the increased peroxynitrite

production (Combs et al., 1999 and 2001). (6) Aβ, in its aggregate form, is capable of

incorporating into planar lipid bilayers by fusion with liposomes and forming cation-

selective channels which are permeable to Ca2+ (Arispe et al., 1994). Aβ-mediated

oxidation on lipids, proteins, and nucleic acids has been extensively studied

(Butterfield and Lauderback 2002). Moreover in vivo studies showed that Aβ induces

and memory deficits, which can be counteracted by treatment with

antioxidants such as idebenone and alpha-tocopherol (Yamada et al., 1999).

17

1.4 Markers for oxidative stress

Researchers investigate oxidative damages in AD brains by using specific markers of oxidized forms of interested molecule species. Numerous markers have been demonstrated in studies of oxidation of lipids, proteins and nucleic acids. Here shows a list of frequently used markers.

Markers for lipid peroxidation: thiobarbituric acid-reactive substances (TBARS) 4-hydroxynonenal (HNE) acrolein lipid-conjugated dienes and trienes, isoprostanes (IsoPs) and isofurans (IsoFs) derived from arachidonic acid neuroprostanes derived from docosahexaenoic acid (DHA)

Markers for protein oxidation (Gracy et al., 1999):

Peptide/amino acid Oxidation products Peptide α-carbon Peptide cross linking, cleavage Cysteine Disulfide, cysteic acid Methionine Met. Sulfoxide, sulfone Tryptophan Hydroxyl-Trp, nitro-Trp, kynurinines Phenylalanine Hydroxyl-Phe Histidine 2-oxoHis, Asn, Asp Arginine Glu semiadaldehyde Lysine Aminoadipic semialdehyde Praline Pyrrolidine, hydroxyPro, pyroGlu, etc. Threonine 2-amino-3 ketobutyric Glutamyl Oxalic acid pyruvic acid Asparagines Aspartic acid, etc. Tyrosine Nitro-tyrosine

18

Marker for reducing sugar oxidation: advanced glycation end products (AGEs),

Marker for oxidation: 5-hydroxyuracil, 5-hydroxycytosine, 8-hydroxyadenine, 4,6-diamino-5- formamidopyrimidine (Fapy-), 8-hydroxyguanine, and 2,6- diamino-4-hydroxy-5-formamidopyrimidine (Fapy-).

1.5 Lipid oxidation in AD

Lipid peroxidation is the mechanism by which lipids, under the attack of ROS,

lose hydrogen atoms from methylene carbons in their side chains. The greater the

number of double bonds in lipids, the easier it is to remove the hydrogen atom

(Pratico and Sung 2004). This is the reason for the high susceptibility of polyunsaturated fatty acids to lipid peroxidation.

Lipid peroxidation is believed to be a prominent form of oxidative damage in

brain due to its high lipid content, especially its high level of polyunsaturated fatty

acids. Lipid peroxides rapidly decompose to liberate reactive carbonyl fragments,

among which malondialdehyde (MDA), 4-hydroxynonenal (HNE), and hexanal are

the major ones. Thiobarbituric acid-reactive substances (TBARS) assay primarily

measures the extent of MDA. Levels of TBARS were elevated in all AD brain

regions except the middle frontal gyrus, and the elevation reached statistical significance in hippocampus and pyriform cortex of AD subjects compared to age- matched controls (Lovell et al., 1995). The level of free 4-HNE was found to be

19

significantly elevated in the ventricular fluid of AD subjects compared to control

subjects (Lovell et al., 1997). An increase of free HNE in multiple brain regions of

AD compared to those of age-matched control subjects was also reported. These

increases reached statistical significance in amygdale, hippocampus and

parahippocampal gyrus, regions showing the most pronounced histopathological

alterations in AD (Markesbery and Lovell 1998). Oxidized products of arachidonic

acid and docosahexaenoic acid (DHA) were also used as indicators of lipid

peroxidation. Arachidonic acid is evenly distributed in gray and white matter,

whereas DHA is particularly enriched in gray matter. Isoprostanes (IsoPs) and

isofurans (IsoFs), derived from arachidonic acid, and neuroprostanes (NeuroPs),

derived from DHA, were shown to be increased in diseased regions of AD brain.

Increased cerebrospinal fluid levels of IsoPs are present in patients with AD early in

the course of their illness (Reich et al., 2001; Montine et al., 2004).

Lipid peroxidation is a deleterious form of oxdative damage because the products from lipid peroxidation have toxic properties to neurons, as proven by many research

groups. 4-HNE is the most frequently studied derivative of lipid oxidation with

respect to AD (Zarkovic 2003). 4-HNE added to rat hippocampus neuronal culture

causes a concentration- and time-dependent decrease in neuron survival (Mark et al.

1997). Incubation of cortical synaptosomes with 4-HNE led to concentration- and

time-dependent changes in membrane protein conformation and bilayer order and motion (Subramaniam et al., 1997). Such kind of changes in protein structure and

20

membrane fluidity can affect the function of the involved membrane proteins. In fact,

4-HNE alters the activities of several membrane transporter proteins. Glutamate transporter GLT1 can be modified by HNE, which impairs glutamate transport activity in synaptosomes and astrocytes. Exposure of rat cortical synaptosomes to 4-

HNE resulted in a decrease of [3H]glutamate uptake (Keller et al., 1997). Impairment of glutamate transport occur in astrocyte cultures after exposure to 4-HNE, and further experiments suggested that HNE promotes intermolecular cross-linking of

GLT-1 monomers to form dimers (Blanc et al., 1998). Accumulation of glutamate in extracellular space may mediate 4-HNE-induced neurotoxicity, which was partially attenuated by NMDA receptor antagonists, suggesting an excitotoxic component to

HNE neurotoxicity (Mark et al., 1997). However, the mechanisms for the neurotoxic

effect of 4-HNE are complex and cumulative. 4-HNE can impair Na+/K+-ATPase

activity and increase neuronal intracellular free Ca2+ concentration. It is known that 4-

HNE can conjugate to specific amino acids or functional groups of proteins and may

alter their functions. Siems et al. (1996) showed that 4-HNE reacts with sulfhydryl groups of Na+/K+-ATPase and inhibits its activity. 4-HNE inhibited mitochondrial respiration by forming stable adducts on mitochondrial proteins (Picklo et al., 1999).

Covalently binding to cysteine, histidine, and lysine residues on synaptosomal

proteins, 4-HNE changes protein conformation and function (butterfield et al., 2002).

Furthermore, 4-HNE can cause cross-linking of tau proteins into high-molecular

weight species that were conjugated with ubiquitin (Montine et al., 1996). HNE

21

binding to proteins introduces carbonyls to proteins, leaving proteins oxidatively

modified as a consequence of lipid peroxidation.

1.6 Protein oxidation in AD

Extensive studies have been performed to investigate the oxidative damage to proteins in aging and Alzheimer’s disease. Oxidation of proteins can lead to the

formation of cross-linked protein aggregates, peptide fragmentation, and conversion

of one amino acid to another or to a modified residue by oxidation of the amino acid

side chain. Of particular significance is the fact that carbonyl derivatives are the

major products of ROS-mediated oxidation of some amino acid residues on proteins

(Amici et al., 1989). Therefore protein carbonyl groups appear to be the most

sensitive marker to measure the level of protein oxidation (Levine et al., 1994).

Hensley and coworkers measured 42 and 37% increase in protein carbonyl content in

AD hippocampus and inferior parietal lobule respectively compared to AD cerebellum, while the level of protein carbonyl content in age-matched normal

controls is comparable in these three brain regions and is relatively low compared to

that in AD corresponding regions (Hensley et al., 1995). Smith et al. (1998) examined

the cellular and subcellular distributions of carbonyl derivatives using in situ 2,4- dinitrophenylhydrazine labeling linked to an antibody system. They detected an increased carbonyl content in NFTs, neuronal cell bodies and apical dendrites in hippocampal sections of AD brains, whereas, in marked contrast, carbonyls were not

found in neurons or glia in age-matched control cases. The increased protein carbonyl

22

formation is not consistently associated with mature senile plaques, which indirectly supports the idea that the soluble form of Aβ aggregates rather than insoluble one induces neurotoxicity (Aksenov et al., 2001). The increased level of protein carbonyl content in some morphologically intact neurons in affected regions of AD brains suggests that oxidative modification of proteins may precede degenerative changes in neurons and is not a simple consequence of cell death.

When superoxide and nitric oxide occur in close proximity, they can spontaneously form peroxynitrite. This highly reactive species can react with proteins at different amino acid side chains depending on the condition. In the absence of CO2, peroxynitrite reacts with methionine, whereas at low levels of CO2, peroxynitrite can

cause tyrosine nitration (Gracy et al., 1999). The leve of nitrotyrosine was elevated

consistently in hippocampus and neocortical regions of AD brains and in ventricular

cerebrospinal fluid (vCSF), reaching quantities five- to eight-fold greater than mean

concentrations in brains and vCSF of age-matched normal subjects (Smith et al., 1997;

Hensley et al., 1998; Tohgi et al., 1999).

For the most cases, there is no repair mechanism for oxidatively modified proteins,

which must be removed by proteolytic degradation. The level of intracellular oxidized

proteins is influenced by two factors: the access of proteins to the attack from ROS

and the degradation rate of oxidized proteins. The accumulation of oxidized forms of

proteins will occur if cells are under oxidative stress or experiencing the down-

23

regulated efficiency of proteolysis. There are five major deleterious consequences

resulting from protein oxidation: (1) some oxidized forms of proteins, such as cross-

linking, can be resistant to degradation by proteasomes and furthermore occupy

proteasomes and impair their ability to degrade other damaged proteins (Grant et al.,

1992; Friguet et al., 1994; Rivett 1986); (2) oxidative modifications can result in the

alteration of protein secondary and tertiary structures and these conformational

changes may expose previously shielded regions to further oxidation or other types of

modifications such as deamidation (Gracy et al., 1999); (3) Oxidized proteins are often functionally inactive (Dean et al., 1997). Oxidation to enzymatic proteins can cause the decrease of their catalytic abilities. Evidence showed that nitration of

tyrosines on manganese SOD and glutamine synthase causes inactivation of those

enzymes (Yamakura et al., 1998; Berlett et al., 1998); (4) structural proteins may also

be the targets of oxidation. Troncoso et al. (1993) demonstrated that oxidation of tau

protein in vitro induces dimerization and polymerization of tau proteins into insoluble

filaments; (5) both carbonylation and nitration can mark proteins for degradation by

the proteinases (Stadtman 1990).

Oxidative modified proteins are localized in the neuronal cytoplasm. The specific

targets of protein carbonyl formation were examined using high-resolution 2D

fingerprinting method and proteomics (Aksenov et al., 2001; Castegna et al., 2002a

and 2002b). Aksenov reported significant increases of specific protein carbonyl levels

in AD brain in β-actin (442 ± 23% of age-matched control) and creatine kinase BB

24

(336 ± 24% of control). Oxidative damage on actin may affect its filament

architecture and cause abnormal assembley of the cytoskeleton (Milzani et al., 1997).

Oxidized CK has less catalytic efficiency leading to decreased availability of ATP in

synaptic terminals. ATP is needed to drive ion-motive pumps, synthesize antioxidant

proteins, and produce other means to inhibit oxidative damage. Carbonyl formation

was detected in severeal other protein species in AD brains by proteomics, including

glutamine synthase, ubiqutin carboxy-terminal L-1, α-enolase, and

dihydropyrimidinase-related protein 2. The profile of nitrated proteins in AD brain

was also screened using proteomics technique. The result showed a highly significant

increase of nitration in AD brain for α-enolase, triosephosphate , and

neuropolypeptide h3 (Castegna et al., 2003). Protein oxidation appears to be a

selective event. The conclusion is based on the fact that oxidation is not observed in

β-tubulin nor glial-derived fibrillary acidic protein (Aksenov et al., 2001).

1.7 DNA oxidation in AD

DNA oxidation by ROS can be very important because DNA is repository of

genetic information present in single copies. The oxidative damage to DNA may play

a role in the pathogenesis of AD. Hydrogen peroxide is a conspicuous species of ROS

in DNA oxidation. Under the diseased conditions, excessive amounts of hydrogen

peroxide escape from antioxidant systems and eventually reach the nucleus due to its

highly diffusible characteristic. Chromatin-bound iron will react with hydrogen

peroxide, producing hydroxyl radicals in situ (Meneghini 1997). This very reactive

25

species will attack a nearby DNA residue causing modifications. ROS-induced DNA damages can be investigated by examining the following events: (1) base modifications; (2) covalent binding of bases within DNA or DNA-protein cross-links;

(3) abasic sites, and (4) strand breaks (Yu and Anderson 1997). The most common form of DNA damage is base modification primarily detected by the technique of gas chromatography/mass spectrometry. Markesbery and coworkers examine nuclear

DNA oxidation in brains of AD patients and age-matched control subjects by measuring oxidized and bases. Statistically significant elevations of oxidized bases were found in AD brains compared to control subjects, and a generally higher level of oxidative DNA damage was present in neocortical regions than that in cerebellum of AD subjects (Gabbita et al., 1998). Studies on ventricular CSF from

AD and normal subjects demonstrated an increase of unrepaired ROS-mediated damage on DNA in AD cases as evidenced by the increased presence of 8-hydroxy-

2'-deoxyguanosine (8-OHdG) in intact DNA and the decreased concentration of the free 8-OHdG that is considered as a repair product (Lovell et al., 1999; Lovell and

Markesbery 2001). The authors suggested that AD brain is subjected to the insults of increased oxidative stress as well as deficiencies in repair mechanisms responsible for removal of oxidized bases from DNA chains. A subsequent study of the activities of enzyme, 8- (responsible for the excision of 8-OHdG), and DNA helicase in nuclear protein samples from the brains of AD and normal subjects was performed by the same group. Their data showed that, compared to the nuclear fractions from the controls, the statistically significant decrease in 8-

26

oxoguanine glycosylase activity was observed in the nuclear fraction of AD-affected regions, while DNA helicase activity was slightly elevated (Lovell et al., 2000). The modest increase in DNA helicase activity in AD may interfere with base excision repair mechanisms. These data indicate that the repair capability for 8-oxoguanine is decreased in AD.

Beal and his colleagues examined oxidative damage to nuclear DNA and mtDNA using 8-OHdG as a marker. They detected a small significant increase in oxidative damage to nuclear DNA and a highly significant three-fold increase in oxidative damage to mtDNA in AD as compared to age-matched controls (Mecocci et al.,

1994). mtDNA is particularly sensitive to oxidative damage with feasible reasons: (1) mitochondria are the major sources of a continuous flux of ROS; (2) mtDNA is close to the inner membrane where ROS are produced; (3) mtDNA is not protected by histone proteins; (4) mitochondria may have a less efficient repair mechanism for

DNA damage than nucleus. Oxidative damages to mtDNA may be responsible for the

AD-related decline of oxidative phosphorylation especially in postmitotic neurons which have high respiratory rates and slow mitochondrial turnover (Linnane et al.,

1989; Tritschler et al., 1993; Wallace 1992). Histone proteins are believed to provide a protective shield for nuclear DNA against oxidative damage mediated by ROS.

However, studies show that histones can be a target for 4-HNE modification because of their high levels of lysine residues. The binding of 4-HNE to histones affects the

27

ability of histones to bind DNA, which increases the vulnerability of DNA to

oxidation in AD brain (Drake et al., 2004).

It is important to study the consequence of DNA oxidation. Oxidatively modified

bases in DNA have been shown to be mutagenic (Cheng et al., 1992). For example,

replication of 8-OHdG as template can cause G → T substitution, and misincorporation of 8-OHdG as substrate can cause A → C substitution. Both events are caused by 8-OHdG. In addition to oxidative modification of bases, strand cleavage is another important DNA lesion under oxidative stress. The DNA strand cleavage can be mediated by the mechanisms involving the attack of hydroxyl radical to the carbon-4 of the sugar moiety or the calcium activation of . Increased levels of DNA breaks were also detected in cerebral cortexes of AD (Mullart et al.,

1990). Multiple mechanisms are utilized to repair DNA damage under physiological conditions. DNA-dependent protein kinase, composed of DNA-PKcs as a catalytic subunit and Ku as a regulatory component, is involved in DNA strand break repair.

Ku DNA binding activity was reduced in extracts of postmortem AD midfrontal cortex compared to that in age-matched controls. Moreover, reduced protein levels of

Ku subunits, DNA-PKcs, and poly(ADP-ribose) polymerase-1 were detected in AD

(Davydov et al., 2003). This change was clearly correlated with the decrease of synaptophysin expression, which is an indicator of synaptic loss.

28

1.8 RNA oxidation in AD.

Compared to oxidation of other molecule species, RNA oxidation in AD is less studied. ROS can hydroxylate purine and pyrimidine bases in RNA into their oxidized forms, among which 8-oxo-7,8-dihydroguanosine (8OHG) demonstrates the highest absolute level and serves as a sensitive biomarker for RNA oxidation (Fiala et al., 1989; Wamer et al., 1997). Abe et al. (2002) reported that the concentration of 8-

OHG in CSF of AD patients was approximately five-fold of that in controls. The concentration of 8-OHG in CSF decreased significantly with the duration of illness and the progression of cognitive dysfunctions, which suggests the increased oxidative

RNA damage in the early stage of the development of AD.

Studies from Dr. Mark Smith’s laboratory (Nunomura et a., 1999) have shown that most of the oxidized nucleosides are associated with cytoplasmic RNA in AD brains. They Figure 1.2. Immunohistochemical method to detect oxidized nucleosides in AD. Oxidized nucleosides are abundant in the neuronal used an in situ approach to identify cytoplasm for a case of AD (A) and it is virtually undetectable in a control (B). 8OHG and found that Nunomura et al. (1999) J Neurosci. 19: 1959- 64. immunoreactivity with the monoclonal antibody 15A3 recognizing both 8OHdG and 8OHG was prominent in the cytoplasm and to a lesser extent in the nucleolus in cases of AD; similar structures, however, were immunolabeled only faintly in controls (Figure 2). Importantly, they also observed that increased RNA oxidation is restricted to vulnerable neurons. RNA

29

oxidation is not dependent on proximity to senile plaques and is reduced in neurons

containing neurofibrillary tangles, suggesting that RNA oxidation may be an early

event preceding cell death. These findings point to the possibility that RNA oxidation

may contribute to neurodegeneration.

1.9 Oxidative stress in other neurodegenerative diseases

Various indices of oxidative damages have been reported within the affected regions that undergo selective neurodegeneration in other neurodegenerative diseases, including Parkinson’s disease (PD), Amyotrphic Lateral Sclerosis (ALS), and

Huntngton’s disease (HD). Markers for lipid peroxidation, protein oxidation, and nucleic acid oxidation have been identified in substantia nigra of patients with PD

(Dexter et al., 1989; Good et al., 1998), spinal motor neurons in ALS (Pedersen et al.,

1998; Beal et al., 1997), and caudate nucleus in HD.

PD: PD is characterized by a selective degeneration of dopaminergic neurons in the substantia nigra pars compacta and the deposition of intracellular inclusion bodies.

The major component of deposits is α-synuclein (Spillantini et al., 1997). Evidence of increased oxidative stress in PD includes: (1) the level of polyunsaturated fatty acids

(an index of the amount of substrate available for lipid peroxidation) was reduced and the level of MDA (a product of lipid oxidation) was increased in parkinsonian substantia nigra as compared to other parkinsonian brain regions and to control tissues (Dexter et al., 1989); (2) extensive accumulations of nitrated proteins including α-synuclein were detected in the inclusions associated with intact and

30

degenerating neurons in substantia nigra of PD (Good et al., 1998; Giasson et al.,

2000); (3) a significant deficiency of GSH was found in the substantia nigra, but not in unaffected brain regions of PD patients (Perry and Yong 1986); (4) studies revealed selective increase of iron in substantia nigra zona compacta of PD patients with the severity of neuropathological changes as compared to age-matched normal controls (Sofic et al., 1991; Gotz et al., 2004).

Much of the interest in association of PD with mitochondrial dysfunction and oxidative stress emerged from studies of 1-methyl-4-phenyl-1,2,3,6- tetrahydropyridine (MPTP)-induced parkinsonism (Tipton and Singer 1993; Bowling and Beal 1995). Monoamine oxidase B (MAO-B) catalyzes the conversion of MPTP to 1-methyl-4-phenylpyridinium (MPP+) which inhibits complex I activity of mitochondria and induces the formation of superoxide radicals (Hasegawa et al.,

1990). Stimulation of inducible nitric oxide synthase (iNOS) and neuronal nitric oxide synthase (nNOS) has been implicated in the MPTP-induced neurotoxic effect

(Klivenyi et al. 2000). After administration of MPTP to mice, there was a robust gliosis in the substantia nigra pars compacta associated with significant upregulation of inducible nitric oxide synthase (iNOS). These changes preceded or paralleled

MPTP-induced dopaminergic neurodegeneration. Mutant mice lacking the iNOS gene were significantly more resistant to MPTP than their wild-type littermates (Liberatore et al., 1999). Specific nNOS inhibitors produced a dose-dependent significant protection against MPTP-induced depletion of dopaminergic neurons in the substantia

31

nigra. NO produced from upregulated NOS reacts with superoxide to form

peroxynitrite which causes nitration of proteins found in PD brain.

Dopaminergic neurons in substantia nigra are the most affected cell group in PD.

The potential of dopamine as a species of oxidant suggests the involvement of

oxidative stress in the pathogenesis of PD. Dopamine, a potential electron donor,

reduces the oxidation state of metals to their reduction state in which metal ions

interact with hydrogen peroxide via Fenton reaction to generate hydroxyl radicals

(Barnham et al., 2004). Neuromelanin, lost in neurons within substantia nigra in PD, was postulated to have antioxidant properties. It protects neurons from dopamine- induced redox-associated toxicity (Smythies 1996; Sulzer et al. 2000). This data suggests that the oxidative stress associated with PD could be the result of a breakdown in the regulation of dopamine/neuromelanin/iron biochemistry.

Identification of mutations linked to familiar PD provides us hints to understand the pathogenesis of the diseases. Mutations on α-synuclein and Parkin are extensively studied. There are two mutations in α-synuclein (A30P and A53T) linked to an autosomal dominant form of Parkinson's disease (Gasser 2001). Both mutations accelerate the aggregation of α-synuclein. It has been suggested that abnormal accumulation of alpha-synuclein can lead to mitochondrial alterations that may result in oxidative stress and, eventually, cell death (Hsu et al., 2000). Another mechanism to associate α-synuclein with oxidative stress is that α-synuclein plays a role in regulating the activity of dopamine. The A53T mutation on α-synuclein impairs vesicular storage of dopamine (Lotharius and Brundin 2002), which results in the

32

accumulation of dopamine in the cytoplasm and subsequent generation of ROS through its interaction with iron. Mutations in Parkin, an ubiquitin protein , are involved in autosomal recessive juvenile parkinsonism. Studies showed that expression of mutant Parkin leads to an elevation of protein carbonyl, lipid peroxide, and nNOS, suggesting that the presence of mutant Parkin induces oxidative stress

(Hyun et al., 2002).

ALS: It is featured by progressive degeneration of the lower motor neurons in the spinal cord and upper motor neurons in the cerebral cortex. There are both familial and sporadic forms of the disease. ALS is characterized by the intracellular deposition of misfolded proteins including Cu/Zn superoxide dismutase (hSOD1). The discovery that a portion of familial ALS results from mutations in hSOD1 gene provides the strongest evidence for a role of oxidative stress in neurodegeneration (Rosen et al,

1993). There are more than 100 mutations of hSOD1 associated with familiar ALS.

Mutations in SOD1 do not impair the activity of enzyme, but cause some toxic gain of function (Gurney et al., 1994). The ROS-generating function of SOD mutants, as measured by the spin trapping method, is enhanced relative to that of the wild-type enzyme (Yim et al., 1996). Proposed mechanisms of the toxic gain-of-function in mutant SOD associated with oxidative stress include: (1) SOD mutants have the decreased Km for hydrogen peroxide (Yim et al., 1996); (2) mutant forms of ALS-

SOD have decreased zinc-binding capacity (Lyons et al., 1996) and removal of zinc from wild-type SOD causes a toxic effect to motor neurons (Estevez et al., 1999;

33

Andersen 2004); (3) inappropriate copper-mediated redox chemistry can be

associated with the toxic effect of mutant SOD as evident by the fact that copper

chelators inhibit the course of the disease in both SOD mutant-cell culture and mouse

models (Wiedau-Pazos et al., 1996; Ghadge et al., 1997; Hottinger et al., 1997).

Although the precise mechanism for the gain-of-function of ALS-mutant SOD is still under investigation and mutations in SOD only account for less than 20% of familial

ALS cases, an increasing body of evidence supports a role of ROS-mediated oxidative stress in ALS. Typical oxidation products such as oxidized proteins, DNA and membrane phospholipids are elevated both in sporadic and familial ALS patients and in several model systems as well (reviewed by Carri et al., 2003). Mitochondrial dysfunction has been reported in ALS postmortem tissues and animal models of ALS

(Mattiazzi et al., 2002; Menzies et al., 2002; Higgins et al., 2003).

Huntington disease (HD): HD is a hereditary neurodegenerative disorder that affects the ability to control movements and induces psychological and cognitive impairments. Caudate nucleus is the most affected brain region in the disease. HD is caused by expansion of an unstable CAG repeat encoding glutamine, close to the 5 end of the chromosome 4 gene for protein (The Huntington’s Disease

Collaborative Research Group 1993). There is evidence that oxidative damage occurs in HD brain and in models of the disorder (Browne et al., 1997; Polidori et la., 1999).

Findings in HD patients include increased incidence of DNA oxidation, protein oxidation and nitration, lipid oxidation (Browne et al. 1999). Evidence suggests the

34

association of mutant huntingtin with energy metabolism defects. Studies showed that mutant huntingtin protein can directly interact with neuronal mitochondria to influence mitochondrial calcium handing (Panov et al. 2003), which may mediate

ROS-induced oxidative damage.

35

CHAPTER 2

QUANTIFICATION OF OXIDIZED mRNA IN ALZHEIMER’S DISEASE

2.1 Abstract

It has been shown that cytoplasmic RNA oxidation occurs to a great extent in the

brains of Alzheimer’s disease (AD) patients. The goal of this study was to investigate whether messenger RNA (mRNA) is oxidatively damaged in AD and to further quantify the magnitude of the mRNA oxidation. A novel immunoprecipitation procedure was developed by our lab to isolate oxidized RNAs using anti-8- oxoguanosine antibody 15A3. Using this approach, we demonstrated that significant amounts of poly(A)+ mRNAs are oxidized in AD brains. Oxidized mRNAs were

separated from non-oxidized mRNAs and quantitative analysis for both mRNA

fractions was performed. Surprisingly, 30 to 70% of the mRNAs isolated from AD

frontal cortices were oxidized while the abundance of oxidized mRNAs was low in

age-matched normal controls. These high abundances implicate the potential contribution of mRNA oxidation to the pathogenesis of AD.

36

2.2 Introduction

Alzheimer’s disease (AD) is characterized clinically as a progressive dementia and pathologically by the accumulation of senile plaques and neurofibrillary tangles in the brain. While the etiology of AD remains largely unknown, oxidative stress is believed to be an important contributor to the disease. Over the past decade, oxidative stress-linked modifications to vital cellular components have been described in association with the vulnerable neurons of AD including lipid peroxidation (Subbarao et al., 1990; Sayre et al., 1997), protein oxidation (Smith et al., 1991; Castegna et al.,

2002a and 2002b), mitochondrial abnormality (Hirai et al., 2001) as well as nucleic acid oxidation (Mecocci et al., 1993; Mecocci et al., 1994). These modifications could lead to impaired cellular functioning and cell death.

ROS hydroxylates guanine to produce 8-oxo-7,8-dihydroguanosine (8OHG) in

RNA (Kasai et al., 1991), which can serve as a sensitive biomarker for oxidative stress (Fiala et al., 1989; Wamer et al., 1997). Several studies have shown an increase of 8OHG level in brains and cerebrospinal fluid of patients with AD (Gabbita et al.,

1998; Lovell and Markesbery, 2001). Nunomura et al. (1999) have used an in situ immunohistochemistry approach to identify oxidized nucleosides in AD. They found that oxidative damage to nucleic acids occurs predominantly in cytoplasmic RNA rather than in nuclear DNA. However, whether oxidative damage to cytoplasmic

RNA contributes to neuronal death in the disease is still unknown. To better understand the involvement of oxidation of cytoplasmic RNA in the pathogenesis of

37

AD, we developed novel procedures to detect and isolate oxidized RNA. We found that messenger RNAs (mRNAs) are oxidized in AD brains. Experiments to determine the magnitude of the mRNA oxidation in brains of AD and age-matched normal controls indicated that 52.3± 6.15% of mRNAs purified from AD frontal cortices were oxidized, while only 1.78±0.56% of mRNAs are oxidized in age-matched normal samples (p<0.001). This study indicates that mRNA oxidation is not a negligible event and may play an important role in the pathogenesis of AD.

2.3 Materials and methods

2.3.1 Tissues

Frozen postmortem brain tissues were obtained at autopsy from 11 AD patients (ages

65-86 years; postmortem intervals 3-11hr), 7 age-matched control subjects

(postmortem intervals 3-18hr), 2 young control subjects (ages 22-49 years; postmortem intervals 6-12 hr) and 7 amyotrophic lateral sclerosis (ALS) patients

(ages 58-65 years; postmortem intervals 6-9 hr). Specimens from hippocampus,

Brodman area 9 and 10 in the frontal cortex and cerebellar cortex were studied.

Among those tissues, 6 AD patients (ages 65-86 years, mean=73.5; postmortem intervals 3-11 hr; 5 male, 1 female) and 6 age-matched control subjects (ages 62-84 years, mean=70.2; postmortem intervals 3-18 hr; 5 male, 1 female) were used for the experiments to the quantification of magnitude of oxidized mRNA in diseased and normal tissues. The CERAD (Consortium to Establish a Registry for Alzheimer’s

Disease) (Mirra et al., 1991) criteria were used for diagnostic categorization of AD.

38

No any other significant gross or microscopic abnormalities were present in these AD

brains. Control subjects had no history of dementia and no evidence of significant neuropathological changes. These postmortem brain tissues were provided by Johns

Hopkins AD Research Center Brain Bank and ALS Brain Bank, Harvard Brain

Tissue Resource Center and University of Maryland Brain and Tissue Banks for

Developmental Disorders.

2.3.2 Isolation of Total RNA

Total RNA was isolated from human postmortem tissues using RNeasy ® Mini

kit (Qiagen, Valencia, CA) according to manufacturer’s direction. Briefly, 100mg

tissue was homogenized in 1ml of Buffer RLT supplemented with 10µl of β- mercaptoethanol (β-ME) by sonication. The lysate was centrifuged at 14,000 rpm for

3 min and supernatant was transferred to a new microtube and used in subsequent steps. One volume of 70% ethanol was then mixed with the cleared lysate, which was applied to RNeasy mini spin column. RNA binding on the column was washed with

buffers RW1 and RPE three times, and eluted into 80µl of DEPC-treated water.

2.3.3 Poly(A)+ RNA Isolation

Poly(A)+ RNA was isolated using the Oligo-tex mRNA Purification kit (Qiagen).

Briefly, 100µg of total RNA was mixed with 250µl buffer OEB and 15µl Oligotex

suspension and incubated at 70˚C for 5 min. The mixture was then incubated in

rotation at room temperature for another 20 min. The mRNA-Oligotex complex was

39

pelleted by centrifugation at 14,000 rpm for 3 min at room temperature. The pellet was washed with buffer OW2 twice and pelleted by centrifugation at 14,000 rpm for

2 min at room temperature. mRNA was eluted off from Oligotex by incubating the pellet with buffer EB at 70˚C for 5 min. After centrifugation at 14,000 rpm for 2 min,

the elution buffer containing mRNA was transferred to a new microtube and the

concentration of mRNA was measured with the OD value at 260nm UV light by

Beckman DU640 spectrophotometry (Beckman Instruments Inc, Palo Alto, CA).

2.3.4 Northern blotting

After DNase I (Invitrogin, Carlsbad, CA) treatment, total RNA and poly(A) +

RNAs were size-fractionated in agarose/formaldehyde gels and blotted onto

positively charged nylon membrane (Roche Applied Science, Indianapolis, IN). The

membrane was hybridized with digoxogenin (DIG)-labeled anti-sense β-actin probe

overnight at 65 oC. The signals were detected with the Digoxogenin High Prime

DNA Labeling and Detection Starter Kit II (Roche Applied Science).

2.3.5 Northwestern blotting

Poly(A) + RNAs were size-fractionated in agarose/formaldehyde gels and blotted

onto positively charged nylon membrane (Roche Applied Science). The membrane

was washed by PBS followed by incubation in 1% BSA blocking solution for 1 hour

at room temperature. The membrane was incubated in a monoclonal anti-8OHG

antibody 15A3 (1:2000; QED Bioscience, San Diego, CA), for 4 hours at room

40

temperature, and then incubated in a secondary peroxidase-conjugated goat anti- mouse antibody (1:1200; ICN, Aurora, OH). The signals were detected with

LumiLight Western Blotting Substrate (Roche Applied Science) according to manufacturer’s directions and exposed on x-ray films (Kodak, Rochester, NY).

2.3.6 Immunoprecipitation of oxidized mRNA

1.5 µg of poly(A) + RNAs were incubated with 2.5 µg of anti-8OHG antibody

15A3 at room temperature for 1 hour. For negative control trials, the primary

antibody was omitted or preincubated with 24 ng/µl 8OHG (Cayman Chemical, Ann

Arbor, MI). 20 µl of immobilized Protein L gel beads (PIERCE, Rockford, IL) were

added and incubated at 4 oC for another 15 hours. The beads were washed three times

by PBS with 0.04 % (v/v) NP-40 (Roche Applied Science). After precipitation of

beads by centrifugation, the oxidized mRNA:antibody:protein L complexes were

separated from non-oxidized mRNAs, which remained in the supernatant. Non-

oxidized mRNAs were precipitated with ethanol and dissolved in 10 µl DEPC-treated

H2O. The following items were added in order into the precipitant of oxidized mRNAs: 300 µl of PBS with 0.04 % NP-40, 30 µl of 10 % (w/v) sodium dodecyl sulfate (SDS), and 300 µl of PCI (phenol: chloroform: isoamyl alcohol; 25:24:1). The mixture was incubated at 37 oC for 15 minutes (vortexing every 5 minutes) and

separated to aqueous phase and organic phase by spinning at 14,000 rpm for 5 minutes. The aqueous layer containing oxidized mRNAs was collected, and mixed with 40 µl of 3 M sodium acetate with pH 5.2, 2 µl of 5 mg/ml glycogen, plus 1 ml of

41

95 % (v/v) ethanol. The sample was frozen at -80 oC for 1 hour and centrifuged for 20

minutes. The pellet was washed by 75% ethanol and air-dried. It was resuspended in

10 µl DEPC-treated H2O.

2.3.7 cDNA synthesis and Southern blotting

Isolated oxidized and non-oxidized mRNAs were reversely transcribed using

AMV reverse transcriptase (Roche Applied Science) respectively. For a 30 µl reaction mixture, 10 µl of mRNAs and 0.75 µg of oligo-(dT) 24-T7 primer were mixed

and incubated at 70 °C for 10 min. After 2 min on ice, a master mix containing 6 µl of

5x first-strand buffer, 0.5 mM each of dATP, dCTP and dGTP, 0.13 mM of dTTP,

0.03 mM of Digoxigenin-11-dUTP (Roche Applied Science), 2.5 U of RNase

Inhibitor (Invitrogen) and 10 U of AMV reverse transcriptase was added. The mixture

was incubated at 42 °C for 90 min. Second strand synthesis was accomplished with E.

coli DNA polymerase I (US Biologicals, Cleveland, OH) in the presence of E. coli

DNA ligase (US Biologicals) and RNase H (US Biologicals) following the manufacturer’s protocol. Double-strand cDNAs were ethanol precipitated and dissolved in 30 µl Tris buffer. 5 µl out of 30 µl digoxigenin-labeled cDNAs were used for further detection by Southern blotting method to compare the difference of cDNA quantities. cDNAs were resolved in 1% agarose gel and then transferred electrophoretically onto positively charged nylon membrane (Roche Applied Science) using the Trans- SD semi-dry transfer system (Biorad, Hercules, CA) according to the manufacturer’s directions. Digoxogenin-labeled on cDNAs were detected with

42

the Digoxogenin High Prime DNA Labeling and Detection Starter Kit II (Roche

Applied Science).

2.3.8 In vitro transcription and oxidation.

RNAs, which were synthesized in vitro, were considered to be non-oxidized.

As the negative control for immunoprecipitation, in vitro synthesized RNAs were

obtained using an amplified antisense RNA (aRNA) technique (Eberwine et al., 1992).

aRNA was made from the double-strand cDNA template containing T7 promotor

using T7 RNA polymerase (US Biologicals). Two rounds of amplification were

performed for the overall 106-fold amplification. For p21ras sequence, the coding

regions plus parts of 5’and 3’ untranslated regions were selected and inserted into the

pcDNA3 vector. PCR was performed by using T7+ and SP6-(dT)30 primers (T7+: 5’-

GGATCCTAATACGACTCACTATAGG; SP6-(dT)30: 5’-

(dT)30CATTTAGGTGACACTATAG). After purification, the amplified cDNAs were transcribed by T7 RNA polymerase (US Biologicals) to RNAs containing poly(A) tail and m7G(5’)ppp(5’)G RNA capping analog (Invitrogen). For oxidation, the capped

poly(A)-tailed RNAs were incubated with H2O2 and cytochrome c (Sigma, St. Louis,

MO) at different dosage at 37 oC for 1 hour. RNA was purified by RNeasy mini kit

(Qiagen).

43

2.3.9 RNA amplification (aRNA)

Immunoprecipitated mRNA and 20 ng of oligo-(dT) 24-T7 primer were mixed with

total volume of 1 µl and incubated at 70 °C for 5 min. After 1 min on ice, a master

mix containing 0.4 µl of 5x first-strand buffer, 0.5 mM dNTP (Invitrogen), 2 U of

RNase Inhibitor and 5 U of AMV reverse transcriptase was added for the first round

of cDNA synthesis. The mixture was incubated at 42 °C for 60 min. Second strand

synthesis was accomplished with 2 U of E. coli DNA polymerase I (US Biologicals,

Cleveland, OH) in the presence of 1 U of E. coli DNA ligase (US Biologicals), 0.2 U

of RNase H (US Biologicals), 0.2 mM dNTP, and 3 µl of 5x second-strand buffer at

15°C for 2 hr. Double-strand (ds) cDNAs were polished with 2 U of T4 DNA

polymerase for additional 15 min, precipitated with etnanol, and dissolved in 5 µl -free water. RNA was amplified from the ds-cDNA by T7 RNA polymerase included in RiboMax In Vitro Transcription System (Promega, Fitchburg, WI).

Amplified RNA (aRNA) and 100 ng of hexamer were mixed with total volume of 5

µl and heat-denatured at 70 °C for 5 min. A master mix containing 2 µl of 5x first- strand buffer, 0.5 mM dNTP, 20 U of RNase Inhibitor and 10 U of AMV reverse transcriptase was added. The reaction was processed at 37°C for 20 min, 42°C for 20

min, 50°C for 10 min, 55°C for 10 min, 65°C for 15 min, and 37°C on hold. 1 U of

RNase H was then added into the reaction and incubated at 37°C for 30 min followed

with 95°C for 2 min. After 1 min on ice, 100 ng of oligo-(dT) 24-T7 primer was added

and incubated at 42°C for 10 min. Then second round of cDNA synthesis was

44

performed with 15 µl of 5x second-strand buffer, 0.2 mM dNTP, and 2 U of E. coli

DNA polymerase I at 15°C for 2 hr. After polishing with 2 U of T4 DNA polymerase, cDNA was transcribed into RNA by T7 RNA polymerase with the same protocol as described above.

2.3.6 Densitometric analysis

The signal intensity from Southern blotting analysis was measured by densitometry using KODAK 1D image analysis software. Comparisons were made between oxidized and non-oxidized mRNA pools of both AD and normal subjects.

2.3.7 Statistical analysis

Data were expressed as mean±S.E.M. and analyzed by Student t-test. P<0.05 were

considered to be statistically significant.

2.4 Results

2.4.1 The quality of RNA from postmortem tissues

The quality of RNA isolated from postmortem brain tissues is always a major

concern of researchers. Before we performed experiments to investigate RNA

oxidation, we evaluated the integrity of total RNA by gel electrophoresis. A good

quality of total RNA was indicated by dense and intact ribosomal RNA bands. The

quality of mRNAs from postmortem tissues was examined by analysis

using a β-actin antisense RNA probe. We evaluated a total of 21 AD and 14 control 45

samples. Only the samples which showed a well-defined single β-actin band,

indicating no degradation, were included in this study. Therefore, 11 AD patients, 7

age-matched control subjects, and 2 young control subjects were carefully selected for further experimental investigation on RNA oxidation.

2.4.2 Detection of oxidized mRNA in AD by Northwestern blotting analysis.

We performed Northwestern analysis to investigate whether poly(A)+ mRNAs

are responsible for RNA oxidation in AD previously observed by Nunomura et al.

(1999). The Northern blot was prepared from poly(A)+ mRNAs isolated from frontal

cortex tissues of AD as well as age-matched normal controls. The blot was probed

with a monoclonal antibody 15A3 that recognizes the oxidized nucleosides 8OHdG

and 8OHG. The immunoreactivity of 15A3 was derived from oxidized RNA in our

experiment because we treated our RNA samples with DNase to avoid DNA

contamination. As shown in Figure 2.1A, the binding of 15A3 to mRNAs was much

more prominent in AD cases than that in control cases. To validate this approach, the

in vitro synthesized p21ras RNAs were used as the negative control and in vitro

oxidized p21ras RNAs, which were prepared by treating in vitro synthesized p21ras

RNAs with H2O2 and cytochrome c, were used as the positive control.

Immunoreactivity with 15A3 antibody was only detected with oxidized ras RNA

(Figure 2.1B). These results directly indicate that poly(A)+ mRNAs are oxidized in

AD.

46

Figure 2.1 Northwestern analysis to visualize oxidized mRNA. (A) Northwestern blot analysis of mRNAs with an anti-8OHG antibody 15A3. mRNAs were prepared from AD or normal frontal cortexes. Significant amounts of mRNAs are oxidized in AD brains. (B) An in vitro generated p21ras RNA (non-ox ras) was used as the negative control and an in vitro oxidized p21ras RNA (ox ras) was used as the positive control. The binding of 15A3 was much more prominent in AD samples than that in control samples.

2.4.3 Immunoprecipitation of oxidized mRNAs

To investigate whether RNA oxidation contributes to the disease, we decided to isolate oxidized mRNA and further determine the magnitude of mRNA oxidation. A novel procedure of immunoprecipitation as shown in Figure 2.2 was developed.

Oxidized mRNAs were separated from non-oxidized mRNAs by immunoprecipitation with 15A3. The immunoprecipitated mRNAs were then

47

Figure 2.2 Immunoprecipitation protocol to isolate oxidized mRNA.

48

reversely transcribed to cDNAs. DIG labeled-dUTPs were incorporated into cDNAs in order to facilitate analysis by Southern blotting.

We applied this immunoprecipitation method to separate oxidized mRNA species from non-oxidized mRNA species. As shown in Figure 2.3A, significant amounts of mRNAs were immunoprecipitated from AD frontal cortex tissues (affected area; lane

3-5) but not from the AD cerebellum (unaffected area; lane 6) or the normal frontal cortex tissues (lane 8-10). A lower amount of mRNAs was immunoprecipitated from

AD hippocampus (lane 7) as compared to the AD frontal cortex. These immunoprecipitated mRNA species are specific to 15A3 as confirmed by the no antibody control (lane 1) and the antibody blocked with 8OHG control (lane 2). In order to rule out the possibility that this antibody may nonspecifically bind to a group of mRNAs which are only present in AD frontal cortex, we synthesized RNAs from the cDNAs that were prepared from the immunoprecipitated mRNAs of AD 1, frontal cortex sample (lane 3, Figure 2.3A). These in vitro synthesized RNAs were then oxidized with H2O2 and cytochrome c. As shown in Figure 2.3B, 15A3 only

precipitated the in vitro oxidized RNAs but not the non-oxidized RNAs. These results

clearly demonstrate that the immunoprecipitated mRNAs are oxidized. We have used

this procedure to analyze 11 AD frontal cortexes, 3 AD hippocampi, 3 AD cerebella

and 9 normal frontal cortexes including presenile and senile tissues. All AD frontal

cortex samples showed strong signals on the Southern blot analyses, moderate signals

for AD hippocampus samples and no or very weak signals for AD cerebellum or

49

normal frontal cortex samples. Among the AD frontal cortex samples, the AD1 sample (Figure 2.3A, lane 3) showed the strongest signal and the AD3 samples (lane

5) showed the weakest signal; signals of the rest samples, such as AD2 (lane 4), were between them. Three AD hippocampal samples showed signals with similar intensities (as represented in lane 7). The postmortem interval did not associate with the amount of immunoprecipitated mRNAs. To exclude the possible role of agonal state in oxidation of mRNAs, we examined the oxidized mRNA levels on amyotrophic lateral sclerosis (ALS) frontal cortexes (n=7) and found no significant amount of oxidized mRNAs (data not shown). These results indicate that significant amounts of poly(A)+ mRNAs are oxidized in AD brain.

50

Figure 2.3 Southern blotting analysis of 15A3-immunoprecipitated oxidized mRNAs. (A) Significantly more oxidized mRNAs were immunoprecipitated from AD frontal cortex (FC) and AD hippocampus (Hip) than normal frontal cortex or AD cerebellum (Cbm). No mRNA was precipitated when the antibody was omitted (no Ab) or pre-absorbed with 8OHG (Ab block). (B) The 15A3 antibody specifically immunoprecipitated oxidized RNAs but not non-oxidized RNAs. The non-oxidized RNA sample (non-ox RNA) was generated from the cDNAs prepared from the immunoprecipitated mRNAs of AD1, frontal cortex sample. The oxidized RNA sample (ox RNA) was treated with 3.2 µM H2O2 and 0.2 µM cytochrome c.

51

2.4.4 Quantification of the level of oxidized mRNAs in diseased and normal tissues

To determine the magnitude of mRNA oxidation, we separated oxidized mRNAs from non-oxidized mRNAs by immunoprecipitation with 15A3 antibodies. Both oxidized and non-oxidized mRNA pools were then reversely transcribed to cDNAs.

DIG- labeled dUTPs were incorporated into cDNAs to facilitate further detection of cDNAs by Southern blotting (Figure 2.4A). We used this procedure to analyze 6 AD frontal cortices and 6 age-matched normal frontal cortices. Signal intensity refers to the amount of DIG-labeled cDNA which reflects the level of mRNAs. In comparison with normal tissues in which the majority of mRNAs were not oxidized, all AD frontal cortex samples had significantly high proportion of oxidized mRNAs (Figure

2.4A). The lowest discrete band in each lane results from the intrinsic products of reverse transcription, which is present in every sample including the no template

control. Densitometric analysis was performed to measure signal intensities for both

oxidized and non-oxidized mRNA pools of AD and normal samples. As shown in

Figure 2.4B, 52.3± 6.2% of mRNAs are oxidized in AD while only 1.78±0.56% of

mRNAs are oxidized in age-matched normal samples (p<0.001).

52

Figure 2.4 Significant amounts of mRNAs are oxidized in AD frontal cortices. (A) Southern blot analysis of oxidized (O) and non-oxidized (N) mRNA pools prepared from AD or normal frontal cortices (n=6 per group). 15A3 antibodies were used to separate oxidized mRNAs from non-oxidized mRNA pools. (B) Densitometric analysis of Southern blot results revealed that 52.3± 6.2% of mRNAs are oxidized in AD while only 1.78±.0.56% of mRNA are oxidized in normal controls. Values are expressed as means±S.E.M.; ** p< 0.001.

53

2.5 Discussion

It has been well documented that nucleic acid oxidation is increased in AD

(Mecocci et al., 1993; Mecocci et al., 1994; Mecocci et al., 1997). Nunomura et al.

(1999) have used a cytological approach to localize oxidized nucleic acids in AD.

They reported that oxidative damage to nucleic acids is restricted to vulnerable neurons and most of oxidized nucleosides are associated with cytoplasmic RNAs. In this study, we employed biochemical approachs to detect the oxidation of mRNA and quantify the magnitude of oxidized mRNA in AD and normal subjects. A monoclonal antibody against 8OHG (15A3) was used to separate oxidized mRNAs from non- oxidized mRNAs. This is the first report of using antibodies to precipitate RNAs. The procedure was validated by using appropriate controls (Figure 2.3). The isolated

oxidized mRNAs are specifically present in cells and not artifacts, as supported by the

following facts: (1) the quality of mRNAs prepared from postmortem tissues was

confirmed by Northern blot analysis using β-actin antisense RNA probe, and only

samples that showed a well-defined single β-actin band, indicating no degradation,

were included in this study, (2) no effect of postmortem intervals on the amount of

oxidized mRNAs was found; (3) no effect of the agonal state in oxidation of mRNAs

was found; (4) in the same AD patient, significant amounts of oxidized mRNAs were

isolated from the affected areas-frontal cortex and hippocampus, but no or very low

amounts of oxidized mRNAs were isolated from the unaffected area-cerebellum.

Hippocampus is the first and most affected brain region in AD. Nunomura et al.

(1999) detected oxidized nucleic acids in healthy-looking pyramidal neurons in AD

54

hippocampus. However in our study less oxidized mRNAs were immunoprecipitated from AD hippocampus as compared to frontal cortex. It is possible that, in AD subjects that we studied, significant numbers of neurons are degenerated so that a less amount of oxidized mRNAs was immunoprecipitated. This also agrees with the study

by Nunomura et al. They demonstrated that nucleic acid damage is not dependent on

proximity to senile plaques and is reduced in neurons containing neurofibrillary

tangles suggesting that RNA oxidation may precede lesion formation.

In this study, we determined the magnitude of mRNA oxidation in AD brains.

We found that 30 to 70% of poly(A)+ mRNAs isolated from AD frontal cortices were oxidized while there was a low proportion of oxidized mRNAs (1.78±0.56%) present

in age-matched normal controls. All poly(A)+ mRNA samples accounted for 3.0-

3.6% of total RNA extracted from each tissue, indicating that the levels of mRNA in diseased and normal tissues were similar. Furthermore, our mRNA extraction procedure did not artificially introduce oxidation of specimens as confirmed by the following observations. First, we extracted RNA from the same tissue using different methods, i.e. TRIZOL (phenol involved) and RNeasy mini kit (no phenol involved), and no difference on the extent of RNA oxidation was found. Second, all tissue samples were processed in a same manner and significant mRNA oxidation was only detected in AD frontal cortex samples but not in normal controls or AD cerebellum

(unaffected area) samples. This indicates that the detected RNA oxidation was not artificially derived from the RNA preparation procedure. Third, we in vitro

55

synthesized RNAs from the cDNAs that were prepared from the isolated oxidized

mRNAs. These RNAs were treated with the same procedures as brain tissues. These

RNAs were not precipitated by 15A3; however, they were precipitated by 15A3 after in vitro oxidization with H2O2 and cytochrome C. This result clearly demonstrates

that the procedure used does not cause artificial oxidation to RNA.

Quantification of the level of oxidized mRNAs was accomplished by reverse

transcribing mRNA to cDNA with incorporation of DIG-labeled dUTPs followed by

Southern blot analysis. Do the modified groups on oxidized mRNAs affect the

efficiency of reverse transcription? We performed reverse transcription of in vitro

oxidized luciferase RNAs and found that the efficiency was not different from the

non-oxidized RNAs. To validate that Southern blots are quantitative when using a

DIG-labeled cDNA, we performed reverse transcription of serial dilutions of mRNAs

prepared from AD brain and found that the signal intensities were proportional to the

amounts of mRNAs, which demonstrated that this method is applicable for

quantitative analysis.

ROS can be generated as metabolic by-products under normal physiological

conditions. Intracellular macromolecules during their lifetimes are under attacks

from ROS in neurons due to their high rate of oxygen consumption. However, below

certain threshold of ROS generation, macromolecular oxidation is minor and is

unlikely to induce pathological changes in neurons. Our result shows that there are

56

1.78±.0.56% of mRNAs oxidized in age-matched normal frontal cortices. Cells can tolerate this level of mRNA oxidation probably by some compensatory mechanisms.

On the other hand, the AD brain is under significant oxidative stress. Our data indicate that about 52.3± 6.2% of mRNAs are oxidized and this may strongly interfere with the fine regulation of cell physiology.

57

CHAPTER 3

CHARATERIZATION OF OXIDIZED mRNA SPECIES IN ALZHEIMER’S DISEASE

3.1 Abstract

Cytoplasmic RNA oxidation is a prominent feature of vulnerable neurons in the brains of Alzheimer’s disease (AD) patients. We previously demonstrated that 30-

70% of messenger RNAs (mRNA) are oxidized in AD brains. The goal of this study was to identify and quantify oxidized mRNA species in AD. We show that mRNA oxidation is not random but highly selective. Importantly, many identified oxidized mRNA species have been implicated in the pathogenesis of AD. Quantitative analysis revealed that some mRNA species are more susceptible to oxidative damage and the relative amounts of oxidized transcripts reach 50 to 70% for some mRNA species.

This high abundance implicates the potential contribution of mRNA oxidation to the pathogenesis of AD.

58

3.2 Introduction

A growing body of evidence implicates that oxidative damage to macromolecules, including lipids (Subbarao et al., 1990), proteins (Smith et al., 1991), and nucleic

acids (Mecocci et al., 1993), are involved in the pathogenesis of Alzheimer’s disease

(AD). Reactive oxygen species (ROS)-induced oxidative damage is restricted to a

defined group of neurons at risk of death in AD.

Nunomura et al. (1999) demonstrated that cytoplasmic RNA oxidation occurs to

a great extent in neurons that are especially vulnerable to degeneration in AD. This

finding raises several interesting questions. The first question is what type of RNAs is oxidized in AD. In our previous study (chapter 2), we demonstrated that messenger

RNA (mRNA) is highly oxidized in AD. Quantitative study showed that 30-70% of mRNAs extracted from AD brains are in their oxidized form. The second question involves the specificity of mRNA oxidation. Which mRNA species are oxidatively

damaged and are there some specific mRNA species more susceptible to oxidative

damage?

In this study, we isolated oxidized mRNAs by immunoprecipitation and

generated cDNA library. The result from cloning analysis showed that specific

functional classes of mRNAs are oxidized and, importantly, many identified oxidized

mRNA species have been implicated in the pathogenesis of AD. The result indicates that mRNA oxidation is not random but selective. Further quantification of

59

individual oxidized mRNA species revealed that some mRNA species are highly

susceptible to oxidative damage as evidenced by the fact that more than 50% of

transcripts are oxidatively modified. This study suggests that oxidative damage to specific mRNA species may play a role in neuronal death in AD.

3.3 Materials and Methods

3.3.1 cDNA amplification and cloning.

For the PCR-based amplification, 2 µl of cDNAs were mixed with 5 µl of 10x

PCR buffer, 3 mM MgCl2, 0.2 mM dNTPs, 1 µg of T7primer, 1 µg of random hexamers and 2.5 U of Taq DNA polymerase (Invitrogen) in a 50 µl reaction volume.

The PCR conditions were 40 cycles of 95 oC for 20 seconds, 40 oC for 3 minutes and

72 oC for 3 minutes. The resulting PCR products were purified using QIAquick PCR

Purification Kit (Qiagen). Afterward, they were digested with MboI (Invitrogen), and

then gel fractioned in 1.0 % agarose gel followed by purification using QIAquick Gel

Extraction Kit (Qiagen) The products were subcloned into pPCR-Script Amp SK (+)

(Stratagene, La Jolla, CA) previously digested with BamH I (Invitrogen) and

dephosphorylated with calf intestinal alkaline (Invitrogen). The vectors

were transformed into SE-SH5α competent cells (Inivtrogen). The plasmid template

DNA was prepared from individual colonies using the Wizard Plus SV Minipreps

System (Promega, Madison, WI), or performing PCR from inoculated culture media

directly.

60

3.3.2 DNA Sequencing and Sequence Analysis

DNA sequencing reaction was conducted using T3 primer at the Plant-Microbe

Genomics Facility of The Ohio State University using the Applied Biosystems 3700

DNA Analyzer with BigDye cycle sequencing terminator chemistry. The acquired sequence data were aligned against the GenBank nucleotide database (National

Center for Biotechnology Information) using BLAST to search for matching sequences.

3.3.3 Filter array and densitometric analysis.

100 ng of PCR product from each sequenced cDNA fragment was used to make filters. For one filter, three sets of PCR products (12 each set) were loaded on 1.0% agarose gel sequentially with 10-minute intervals and resolved by electrophoresis

(100 V). After transfer of linearized PCR-amplified cDNAs to nylon membranes (i.e. filters), digoxigenin-labeled cDNAs from oxidized and non-oxidized mRNA pools of the frontal cortex or cerebellar cortex of AD brain and control brain were used as probes to hybridize to the filters for 20 hours at 42 oC. The signals were detected by

DIG High Prime DNA Labeling and Detection Starter Kit II (Roche Applied Science)

and exposed on x-ray films for 4 hours, 12 hr ,and 24 hours. The signal intensity of

each band was measured by densitometry using KODAK 1D image analysis software.

The mean density of a given cDNA among the AD cases or control cases was

calculated and significant differences between AD and control groups were evaluated.

61

3.3.4 Primers used for Polymerase Chain Reaction.

With gene-specific primers, polymerase chain reaction (PCR) was performed on

cDNAs and PCR-amplified cDNA products. The following primers specific for human gene transcripts were used: cytochrome b F (5’-gctactcagtagacagtcccaccc-3’),

R (5’-gaggaggtctgcggctaggagtc-3’); apolipoprotein D F (5’-agatggtgatgctgctgct-3’), R

(5’-gcagcctccctgtagaacctgg-3’); HRAS1 F (5’-atgacggaatataagctggtggtggtgg-3’), R

(5’-tcaggagagcacacacttgcagct-3’); β-actin F (5’-cgggacctgacagactacctcat-3’), R (5’- accgactgctgtcaccttcacc-3’); EAAT2 F (5’-ggcaactggggatgtaca-3’), R (5’- acgctggggagtttattcaagaat-3’); neurofilament light chain F (5’- ggacaagcagaacgccgacatca-3’), R (5’-cttgttcctccccagcaccttca-3’); APP F (5’- tgaacgtgggagttcagctgcttc-3’), R (5’-aaagaagggtttgtccaggcatgc-3’); tau F (5’- cctgctgtgggtcagtgtgcc-3’), R (5’-aggggtataggcagtgattgggct-3’). PCR products were visualized on 1% agarose gels stained with ethidium bromide.

3.3.5 Semi-quantitative RT-PCR analysis

With gene-specific primers, polymerase chain reactions (PCRs) were performed on cDNAs derived from both oxidized and non-oxidized mRNAs. The following primers specific for human gene transcripts were used: Cu/Zn SOD1 F (5’- atggcgacgaaggccgtgtgcgt), R (ctggcaaaatacaggtcattgaaacagaca); carbonyl reductase 1

F (5’-ggagctggtggggctcatgaac), R (5’-agtcaactgagtgcgtaggttgc); cytochrome b F (5’- gctactcagtagacagtcccaccc-3’), R (5’-gaggaggtctgcggctaggagtc-3’); presenilin 1 F (5’- gggcaggggttccagcttcc-3’), R (5’-acagcagggctgacagcacc-3’); apolipoprotein D F (5’-

62

agatggtgatgctgctgct-3’), R (5’-gcagcctccctgtagaacctgg-3’); transferrin F (5’-

ccaacaacaaagagggatactacgg), R (5’-cagaaggtccttggtttccga-3’); α-Enolase 1 F (5’-

ggagactgaagataccttcatcgc), R (5’-gaagttctaaggaagcggtacgaa); Calpain, small subunit 1

F (5’-aaacagttcgacactgaccgatc), R (5’-caggtacaggggagaggttacg); cadherin 18, type 2 F

(5’-tttgagtacaggagccttaatcgc), R (5’-agtcataagggggaacgctagg); Glutamate

dehydrogenase 1 F (agttccaagacaggatatcgggt), R (5’-actcagcttgtacatgatccgcta). PCRs

were performed in the presence of 3 mM MgCl2, 0.2 mM dNTP, 0.25 µM primers and 2 U Taq DNA polymerase (Invitrogen) in 1x PCR buffer. In order to verify that amplification of each gene was within the linear range, a series of cycles (20, 25, 30 cylces) were performed (95 °C for 30 s, specific annealing temperature to each set of primers for 45 s, and 72 °C for 1 min). PCR products were visualized as single bands

on 1% agarose gels stained with ethidium bromide. The band densities were

measured by densitometry using KODAK 1D image analysis software. Comparisons

were made between oxidized and non-oxidized mRNA pools of both AD and normal

cases.

3.3.6 Statistical analysis

Data were expressed as mean±S.E.M. and analyzed by student t-test. P<0.05 were

considered to be statistically significant.

63

3.4 Results

3.4.1 Molecular cloning and sequence analysis of oxidized mRNAs from Alzheimer’s

brains.

In the previous study (chapter 2), we isolated oxidized mRNAs from non-

oxidized mRNAs by immunoprecipitation using 15A3 antibody. To identify oxidized

mRNA species in AD, we cloned the cDNAs that were prepared from the

immunoprecipitated mRNAs of AD1, frontal cortex sample (Figure 2.2A, lane 3) into pPCR-Script Amp SK(+) vector. Two procedures were applied for efficient cloning. We first amplified cDNAs using PCR method. As shown in Fig 3.1, the abundance of PCR-amplified products from AD cDNA library was much higher than that from age-matched controls. We digested the amplified cDNAs with the four- cutter enzyme MboI. The MboI-digested fragments, which represent each oxidized mRNA species, were efficiently cloned into the vector. A total of 340 clones were

randomly selected for further analysis. They were first analyzed by HaeIII restriction

. The clones with different restriction patterns were subjected to sequencing.

63 different transcripts were consequently identified. They were grouped into 10 categories as shown in Table 1: 24 transcripts are known genes; 17 correspond to uncharacterized genes; 12 correspond to uncharacterized EST clones; 10 are

unidentified genes. For those known genes, their gene products are involved in free

radical modulation, detoxification, cell death pathway, signal transduction, synaptic

plasticity, long-term potentiation and cell proliferation. Importantly, at least 10 of

those known genes have been previously characterized in AD, which include p21ras,

64

MAPK kinase 1, α-enolase 1, carbonyl reductase, apolipoprotein D, transferrin, phosphotriesterase protein, glutamate dehydrogenase, calpain and Cu/Zn superoxide dismutase (SOD1).

Figure 3.1 Electrophoresis of PCR-amplified cDNA products. Lane 1 is molecular weight marker of DNA. Lane 2 and 3 are cDNA amplification products from immunoprecipitated mRNA in the frontal cortexes of age-matched normal control and AD subjects.

65

Description Genbank Ratio accession no. intracellular signal transduction and intercellular communication tubby super family protein NM_020245 1.0 mammary tumor and squamous cell carcinoma- NM_138565 1.9 ± 0.10 associated (p80/85 src substrate) (EMS1), transcript variant 2 cell death-inducing DFFA-like effector b AF218586 3.0 ± 0.23 purinergic receptor P2X U83993 6.0 ± 0.32 MAPK kinase 1 L11284 8.7 ± 0.96 cadherin 18, type 2 XM_011214 8.8 ± 0.30 c-Ha-ras J00277 10.3 ± 0.56 gene regulatory proteins

SWI/SNF related matrix associated, actin dependent XM_009081 3.1 ± 0.69 regulator of chromatin subfamily a, member 4 inhibitor of DNA binding, dominant negative helix- XM_046179 4.0 ± 1.10 loop-helix protein helicase/primase complex protein AF487338 6.0 ± 1.25 ribosomal protein L7a BE559788 6.3 ± 0.67 alpha enolase 1 NM_001428 9.9 ± 1.06 free radical modulation proteins carbonyl reductase 1 XM_009749 4.5 ± 1.25 Cu/Zn SOD1 X02317 6.8 ± 1.10

Table 1 continued

Table 1 Classification of Clones and statistical analysis on filter array. The ratio represents the relative abundance of each oxidized RNA species in AD frontal cortex and refers to the fold change by normalizing the value of each band to the least intense band (i.e. tubby super family).

66

continued

Intra- or intercellular transport proteins

apolipoprotein D XM_049984 2.1 ± 0.25 transferrin XM_002793 7.5 ± 1.46 phosphotidylinositol transfer protein beta NM_012399 9.6 ± 0.86

Mitochondrial related

cytochrome b AF254896 7.1 ± 0.68 mRNA containing 12s rRNA sequence AY012136 17.3 ± 3.00

Metabolism-related

phosphotriesterase related AI819400 2.3 ± 0.17 uridine-cytidine kinase 1 (UCK1) XM_033385 2.7 ± 0.54 glutamate dehydrogenase 1 NM_005271 5.9 ± 0.68 Calpain, small subunit 1 XM_009338 18.2 ± 1.50

Cytoskeleton regulation

RNB6 XM_007244 2.9 ± 0.46

Uncharacterized mRNAs with complete open reading frame

Clone image: 5261941 BC035097 1.5 ± 0.23 FLJ10743 NM_018201 1.9 ± 0.51 FLJ12123 AK022185 1.9 ± 0.19 FLJ10994 AK001856 2.4 ± 0.60 LOC257167 XM_173054 2.6 ± 1.15 FLJ12387 similar to kinase light chain BC034373 3.1 ± 0.31 FLJ14220 AK024282 5.1 ± 1.14 FLJ12936 AK022998 3.5 ± 0.66 FLJ 14936 NM_032864 4.4 ± 0.17 FLJ31340 AK055902 5.4 ± 1.22 MAC3077 XM_018160 5.7 ± 0.51

Table 1 continued

67

continued FLJ33905 AK091224 6.1 ± 1.06 Similar to FK506 binding protein 9 BC007443 6.5 ± 1.32 FLJ13658 AK023720 6.5 ± 0.87 Secreted protein of unknown function BC008823 8.5 ± 1.45 AD-003 protein XM_052723 14.4 ± 0.74 Clone MGC: 33424, IMAGE: 5266665 BC036526 30.9 ± 2.40

EST clones

EST BI463414 2.9 ± 0.49 EST BM706879 3.1 ± 1.07 EST AA584851 4.1 ± 0.35 EST AA084569 6.7 ± 2.09 EST AV755263 7.6 ± 1.58 EST BG013057 8.7 ± 0.96 EST AI457597 9.9 ± 0.81 EST AI720416 10.3 ± 1.64 EST AA828749 11.2 ± 2.30 EST AA828970 11.2 ± 1.99 EST AA923791 21.8 ± 1.71 EST AA488746 27.5 ± 2.43

Unidentified

1 1.7 ± 0.47 2 3.0 ± 0.84 3 5.0 ± 0.80 4 5.2 ± 0.58 5 6.1 ± 0.40 6 10.2 ± 1.65 7 13.7 ± 2.31 8 20.8 ± 2.17

Table 1 continued

68

continued 9 21.0 ± 3.43 10 22.7 ± 1.31

To exclude the possibility that some immunoprecipitated mRNAs were derived

from nonspecific binding to the antibody or protein L, we examined some

immunoprecipitated mRNA species, i.e. apolipoprotein D, cytochrome b and p21ras

mRNAs, by RT-PCR in the no antibody control, in which no 15A3 antibody was

added in immunoprecipitation (see Figure 2.2A lane 1), or the antibody block control,

in which 15A3 antibody was preabsorbed with 8-OHG. The results showed that they

were not present in both controls (Figure 3.2).

Figure 3.2 PCR analysis of identified clones. Some immunoprecipitated mRNA species from AD frontal cortex (FC) were examined in experimental control trials by PCR. Omission of antibody in the immunoprecipitation procedure (lane 2: AD1, FC, no Ab) or preabsorption of antibody with 8-OHG (lane 3: AD1, FC, Ab block) yielded no signal of PCR-amplified cDNAs including apolipoprotein D (apo D), cytochrome B (cyto B), and p21ras (ras), which were present in oxidized mRNA pool of AD frontal cortex as shown in lane 4. Lane 1 is the molecular weight marker of DNA.

69

3.4.2 Filter array to examine the susceptibility of individual mRNA species to oxidation in AD.

We employed a filter array method to quantitate the level of oxidation for each identified mRNA species and also to screen these oxidized mRNA species in other

AD and normal control cases. The filters were blotted with the equal amount of each identified cDNA fragments and then probed with a DIG-labeled cDNA probe that was prepared from the immunoprecipitated mRNAs (Figure 3.3). All samples were processed in the same manner. The signal intensity of each band represents the abundance of the corresponding mRNA species in the oxidized mRNA pool. It clearly demonstrates that the identified mRNA species are highly oxidized in the AD frontal cortex (AD3, FC, 4hr) and much less oxidized in the normal frontal cortex

(Normal 1, FC, 4hr) or the AD cerebellum (AD3, Cbm, o/n). Importantly, these results also indicate that some mRNA species are more susceptible to oxidative damage as shown by the stronger signal in the filter array. Furthermore, it appears that most of mRNA species that show oxidative damage in AD frontal cortex are also oxidized in the normal frontal cortex but to a much lesser extent. These signals are specific to the identified mRNA species because no signal was detected in negative controls in which 15A3 was blocked by 8OHG (AD3, FC, Ab block, o/n) or was omitted during the immunoprecipitation. Densitometric analysis was performed for each oxidized mRNA species in AD brains (n=11) and control brains (n=9) on filter arrays (Table 1). Normalization of the signal intensity from a desired band to that from the least intense cDNA band (tubby super family protein) across all filter arrays

70

enables a reliable indicator for the relative abundance of oxidized form of each oxidized mRNA species. All 11 AD samples had the same pattern of signal intensities on the filter arrays, indicating that the oxidized RNA profile is similar amongst AD

patients.

Figure 3.3 Filter array analyses of identified oxidized mRNA species reveal that some mRNA species are more susceptible to oxidative damage to others. DIG-labeled cDNAs prepared from the immunoprecipitated mRNAs were used as probes to hybridize DNA fragments, which were PCR-amplified from the identified clones and then immobilized onto nylon membranes. The signal intensity of each band represents the abundance of the corresponding mRNA species in the oxidized mRNA pool. The identified mRNA species are highly oxidized in the AD frontal cortex (AD3, FC, 4 hr exposure to film) and much less oxidized in the normal frontal cortex (Normal 1, FC, 4 hr exposure) or the AD cerebellum (AD3, Cbm, o/n, overnight exposure). No signal was detected in the negative controls in which 15A3 was blocked by 8OHG (AD3, FC, Ab block, o/n) or was omitted during the immunoprecipitation (not shown).

71

3.4.3 Quantification of the magnitude of oxidation for individual mRNA species by

filter array and semi-quantitative RT-PCR.

In the previous section (3.4.2), we measured the relative abundance of oxidized

form for individual mRNA species. The result showed that some mRNA species are

highly susceptible to oxidative damage. However, the intracellular abundance of

transcripts varies in different mRNA species. Therefore, the severity of oxidation can

be expressed by the percentage of oxidized transcripts for identified mRNA species.

We first applied filter array to address this issue. Two duplicated filters were made by

immobilizing cDNA fragments, corresponding to the identified oxidized mRNA

species, onto nylon positively-charged membranes. One filter was hybridized with a

probe prepared from oxidized mRNA pool of AD or normal control subjects. The

other was hybridized with a probe prepared from non-oxidized mRNA pool. Both

filters were processed in the same manner. The signal intensity of each band

represents the abundance of that corresponding mRNA species in each pool (Figure

3.4 A). All AD samples yielded signals on both filters, while the normal control

samples had signals on the filter hybridized with the non-oxidized mRNA probe and

did not consistently show signals on the filter hybridized with the oxidized mRNA

probe, which make the statistical analysis of normal samples inapplicable.

Densitometric analysis was performed to quantify the levels of oxidized RNA versus non-oxidized RNA for each mRNA species in AD samples, which gave detectable signals. We also applied semi-quantitative RT-PCR to analyze some mRNA species including Cu/Zn SOD1, carbonyl reductase 1, cytochrome b, presenilin 1,

72

apolipoprotien D, transferrin, α-enolase 1, calpain small subunit 1, cadherin 18 type 2,

and glutamate dehydrogenase 1 (some shown in Figure 3.4 B). As summarized in

Table 2, some mRNA species such as cytochrome b and glutamate dehydrogenase 1

are highly oxidized in AD frontal cortices (n=6).

Figure 3.4 Quantification of the magnitude of oxidation for individual mRNA species by filter array and semi-quantitative RT-PCR. Two duplicate filters were probed by DIG- labeled cDNAs from non-oxidized and oxidized mRNA pools respectively. The signal intensity of each band represents the abundance of that corresponding mRNA species in each pool (A). Semi-quantitative RT-PCR was performed with different cycle numbers (25 and 30) in non-oxidized (N) and oxidized (O) mRNA pools (B).

73

mRNA species Oxidized mRNA/total mRNA (%)

Cytochrome b 75.1±8.4 Calpain, small subunit 1 73.1±10.7 Glutamate dehydrogenase 1 71.9±11.7 Transferrin 66.7±2.9 Cu/Zn SOD1 66.6±3.3 α-Enolase 1 65.6±6.0 Apolipoprotein D 64.9±2.9 Similar to FK506 binding protein 9 63.4±5.8 Carbonyl reductase 1 62.1±3.1 cadherin 18, type 2 58.3±8.2 EST, AA828749 62.1±4.0 EST, AI457597 60.3±4.2 EST, AA828970 58.4±5.2 EST, AI720416 57.8±2.7 EST, AI141130 57.7±1.8 EST, AA923791 57.1±4.0 EST, BE043705 56.1±1.7 EST, AA488746 54.8±1.2 EST, AA084569 53.9±2.2

Table 2. The magnitude of oxidation of specific mRNAs in AD frontal cortices. The values are expressed as means±S.E.M.

74

3.4.4 Characterization of the selectivity of oxidation to individual mRNA species.

Whether mRNAs are randomly attacked by ROS or, for some unknown reason, specific groups of mRNA are oxidized is an interesting question. Our cloning analysis did not identify some highly abundant mRNA species, such as β-actin, in oxidized mRNA pool of AD. To avoid the possibility that we lost some mRNA species in the processing of cloning procedure, we used RT-PCR to examine whether highly abundant mRNA species (GAPDH, β-actin), neuron-specific mRNA species

(neurofilament light chain protein), astrocyte-specific mRNA species (glutamate transporter EAAT2), and some AD-related genes (amyloid precursor protein and tau) are present in the oxidized mRNA pool of AD. All postmortem tissues included in this study were examined. Interestingly, no β-actin, neurofilament light chain protein, or EAAT2 mRNA was found in the oxidized portion of any AD or normal control samples. Further, the APP and tau mRNAs were also not found in the oxidized mRNA pool. The presenilin 1 mRNA was found in two AD frontal cortex samples that had high levels of oxidized mRNA. These results indicate that RNA oxidation is not random but highly selective.

3.5 Discussion

In our previous study (chapter 2), we showed that 52.3± 6.2% of mRNAs are oxidized in AD, while 1.78±.0.56% of mRNAs oxidized in age-matched normal subjects. This high value indicates that mRNA oxidation is not a negligible event and may play an important role in the pathogenesis of AD. To better understand the

75

contribution of mRNA oxidation to the disease, we investigated: (1) what mRNA

species are oxidized in AD; (2) how severe the individual mRNA species is oxidized

in AD; (3) whether mRNA oxidation is a random or selective event.

In this study, we used a cloning strategy to characterize oxidized mRNA species

in AD. 63 different transcripts were identified. 24 transcripts are known genes. The

very striking finding is that most of the identified known oxidized transcripts are related to AD. Either these transcripts or their family members have been characterized in AD or their protein functions have been implicated in the pathogenesis of AD. The p21ras proteins as well as down-stream elements of the

MAPK cascade, including MAPK kinase 1, are elevated at very early stages of AD

(Arendt et al., 1995; Gartner et al., 1999). The increased expression of these proteins is associated with neurofibrillary degeneration. Furthermore, it has been shown that

p21ras/MEK/ERK pathway modulates the expression and posttranslational

processing of tau and APP proteins (Ferrer et al., 2001; Mills et al., 1997; Sadot et al.,

1998). Carbonyl reductase, which plays an important role in detoxification of protein

carbonylation, was found increased in several brain regions of AD (Balcz et al., 2001).

Superoxide dismutase (SOD1), an antioxidant enzyme, was found to have increased

expression in AD brains, but its activity was reduced (Omar, 1999).

Phosphotriesterases catalyze the hydrolysis of a range of phosphotriester compounds.

Paraoxonase, a member of phosphotriesterase family, can metabolize oxidized lipids thereby functioning as an antioxidative enzyme (Furlong, 2000). activity

76

is significantly lower in AD than that in healthy controls (Paragh et al., 2002).

Glutamate dehydrogenase plays an important role in maintaining extracellular glutamate level and its activity is decreased in lymphocytes from AD patients

(Iwatsuji et al., 1989). Cytochrome b is a subunit of complex III of the mitochondrial respiratory chain. A mutation in cytochrome b causes oxidative stress and aging in nematodes (Ishii et al., 1998). Several point mutations of cytochrome b have been associated with Parkinson disease (Fisher & Meunier, 2001; Tanaka, 2002; von

Eitzen et al., 2000). Apolipoprotein D (apo D), a member of the lipocalin family, is increased in cerebrospinal fluid and hippocampus of AD (Terrisse et al., 1998; Belloir et al., 2001). Transferrin, an iron-transporting protein, is significantly increased in AD brains (Focht et al, 1997; Loeffler et al., 1995). The transferrin C2 variant is associated with late-onset AD (Namekata et al., 1997; van Rensburg et al., 2000).

Purinergic receptors are involved in extracellular nucleotide-mediated fast excitatory synaptic transmission. Cadherins, a family of cell adhesion molecules, play important roles in synaptic structure, function, and plasticity (Goda, 2002). These molecules have been implicated in the induction of long-term potentiation (LTP) of hippocampal synaptic strength, a cellular model for learning and memory (Huntley et al., 2002). RNB6 is a member of Ena/VASP family that is involved in actin filament assembly (Ohta et al., 1997). The Ena/VASP family protein has been suggested to contribute to the perpetuation of the plastic changes in synaptic transmission during

LTP by regulating actin polymerization (Kato et al., 1997). Calpains, a family of calcium-activated cysteinyl/thiol proteases, have been implicated in

77

neurodegenerative processes (Chan and Mattson, 1999). Calpain activation is increased in AD and the active form of calpain 2 is colocalized with hyperphosphorylated tau protein (Adamec et al., 2002; Nixon, 2000). Cell death- inducing DFF45-like effector B (CIDE-B) shares homology with the apoptosis- inducing DNA fragmentation factor (Inohara et al., 1998). In summary, the mRNA species that are heavily oxidized in AD frontal cortex normally serve a variety of important physiological functions. Although these gene transcripts and their protein products have not been established to be very important in AD, their roles in AD may be underestimated and need further investigations. Furthermore, oxidation of one mRNA species may not affect normal cellular function; however, oxidation of numerous mRNA species may strongly interfere with the fine regulation of cell physiology that may contribute to the pathogenesis of AD. One interesting observation is that most of these AD-related mRNA species have been reported upregulated in AD. It is conceivable that oxidized mRNAs may produce defective proteins and the cells have to compensate for the loss of functional proteins by increasing their expressions. However, the increased mRNAs may be still damaged and produce defective proteins, resulting in loss of activity. Furthermore, the defective proteins may have not only a loss of normal function but also a gain of toxic function.

One important finding in this study is that mRNA oxidation is not random but highly selective and some mRNA species are more susceptible to oxidative damage.

78

The highly abundant mRNAs such as β-actin mRNAs are not oxidized in AD. Those

identified known oxidized mRNA species are not abundant in cells. This would indicate that mRNAs are not randomly hit by free radicals. RNA oxidation may be regulated by unknown mechanisms. We have examined RNA sequences and RNA structures (http://www.tbi.univie.ac.at) for those identified mRNA species and no common motifs or structures were found. The spatial conformation of mRNAs may regulate the rate of base hydroxylation by oxidants. mRNA-binding proteins can protect transcripts from the oxidative damage. RNA stability may affect the possibility of individual mRNA species to be oxidized under oxidative attacks. Also the turnover rate of oxidative damaged mRNAs may be a factor to determine their presence in the cytosol. Studies from Butterfield group on oxidatively modified proteins reveal the presence of specific targets of protein oxidation in AD brain

(Castegna et al., 2002a, 2002b). Taken together, these observations on protein and

RNA oxidation may indicate that specific macromolecules are more prone to free radical oxidation in AD.

This is the first study to characterize oxidized mRNA species in AD, which contributes to a growing body of evidence of oxidative damage in AD. We quantified the degree of oxidation for individual mRNA species in AD (Table 2). Those mRNA species with consistently high values in different tissues, meaning significance in statistic analysis, are listed in the table. Some mRNA species are highly oxidized as shown by the fact that the majority of transcripts are in their oxidized form. This

79

highly significant modification is likely to be of serious consequences, which should not be underestimated. Further investigation on the consequence of mRNA oxidation is of critical importance for a better understanding on how oxidized mRNAs are involved in the pathogenesis of AD.

80

CHAPTER 4

THE CONSEQUENCE OF mRNA OXIDATION

4.1 Abstract

We have previously demonstrated that 30-70% of mRNAs are oxidized in affected brain regions of Alzheimer’s disease (AD). Some specific mRNA species are highly susceptible to reactive oxygen species (ROS)-induced oxidative damage and the majority of transcripts of those species are in their oxidized form. This suggests that mRNA oxidation may be involved in the pathogenesis of AD. The studies on the functional effects of oxidized mRNA would provide us insights on how mRNA oxidation is associated with AD. We have investigated this issue by expressing oxidized mRNA in an in vitro translational system and cell lines. Results showed that mRNA oxidation can cause the decrease of normal protein expression as well as protein activity. The abnormal loading of polyribosome complexes on oxidized mRNA transcripts was detected in the process of translation. To investigate how cells function with different amounts of intracellular oxidized mRNAs, we injected oxidized mRNAs to a hippocampal neuronal cell line. The data showed that microinjection of oxidized RNA into cells can lead to cell death when the

81

concentration of injected RNA reaches the certain level. These results indicate the possibility of oxidized mRNA to interfere with physiological functions of cells and

implicate the potential contribution of RNA oxidation in the pathogenesis of AD.

4.2 Introduction

Increasing evidence supports the role of oxidative modification of nucleic acids

in neurodegenerative diseases, including Alzheimer’s disease (AD) (Mecocci et al.,

1994; Nunomura et al., 1999; Shan et al., 2003), Parkinson’s disease (PD) (Zhang et

al., 1999; Kikuchi et al., 2002), and Amyotrophic Lateral Sclerosis (ALS) (Chang et

al., 2004). Localization of oxidized nucleotide in cytoplasm of vulnerable neurons in

AD indicates that RNA is the main target of oxidation damage. It is known that RNAs

require Mg2+ for proper folding and activity. Redox-active iron (Fe2+) has been shown

to be capable of interacting more strongly at metal binding sites of RNA than Mg2+.

Fe2+, which binds on RNA, is oxidized in the presence of hydrogen peroxide to

generate highly reactive hydroxyl radicals by the Fenton reaction (Berens et al., 1998),

which causes RNA oxidation in their proximity. Altered iron homeostasis has been

found in AD. An aberrant accumulation of iron was detected in cytoplasm of

vulnerable neurons in AD brain (Smith et al., 1997), which was suggested to be the

source of reactive oxygen species (ROS) triggering oxidative damage to

macromolecules in AD.

82

The consequence of RNA oxidation remains largely unexplored. Studies of retrovirus, in which RNA serves as the genomic material, demonstrated that the potential of mutagenesis exists as a result of base misincorporation opposite to the oxidized bases on RNA (Rhee et al., 1995). This suggests that oxidatively modified

RNA can interfere with correct base pairing, which may compromise the accurary of cellular processes such as protein translation in eukaryocytes. More hints can be extracted from the studies of DNA oxidation, which has been more extensively studied as compared to RNA oxidation. DNA oxidation can be mediated by the

Fenton reaction in various conditions (Luo et al., 1996; Henle et al., 1999).

Hydroxylation of nucleotides in DNA and RNA is one of the principal modifications by ROS (Warmer et al., 1997). Studies on the consequence of DNA oxidation demonstrated that oxidized bases are mutagenic (Cheng et al. 1992). During replication, 8-OHdG can pair with all three other bases rather than just C. Moreover,

8-OHdG present in the free nucleotide pool can be misincorporated as substrate during DNA replication, therefore causing the substitution of A to C. DNA damage may lead to signal cell-cycle arrest and programmed cell death (Berardi et al., 2004) and a point mutation in p21ras gene mediated by 8-OHdG may change cell destiny.

Cells develop multiple mechanisms to repair DNA damage under physiological conditions. RNAs, unlike with two or four copies per cell, are present at tens to thousands of copies and are readily replaced. However, damages to RNAs should not be underestimated. Abnormal transcripts of mRNAs have been proved to be capable of causing neurodegeneration directly without the involvement of corresponding

83

proteins (Jin et al., 2003). Furthermore, the presence of RNA repair mechanism, which was discovered recently by Aas et al. (2003), indicates a greater investment in the protection of RNA than previously suspected. Damaged RNA could impair normal cell functions on its own or via some modification of protein translation.

In our previous study (chapter 2), we found that 30-70% of mRNAs are oxidized in AD brain. Oxidatively damaged mRNAs in such a high abundance may act synergistically in easing neurodegeneration in AD. Here we examined the consequence of mRNA oxidation using an in vitro translational system and cell lines.

The results showed that RNA oxidation can lead to the impairment of protein translation as well as cell death. These findings support the role of mRNA oxidation in the pathogenesis of AD.

4.3 Materials and Methods

4.3.1 Cell cultures

HEK293 cells were obtained from American Type Culture Collection (ATCC,

Manassas, VA) and were cultured in Minimum Essential Medium with Eagle’s salts

(MEM; Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS;

Invitrogen), 100 µM MEM nonessential amino acids (Invitrogen) and penicillin- streptomycin-glutamine (100 U/ml-100 µg/ml-2 mM; Invitrogen). Neuro2A cells were obtained from ATCC and PC12 cells were generously provided by Dr. Richard

W. Burry (Department of Neuroscience). They were maintained in Dulbecco’s

84

modified Eagle’s medium (DMEM; Invitrogen) supplemented with 10% FBS and

penicillin-streptomycin-glutamine.

4.3.2 Plasmid DNA contruction

The complete open reading frames of hSOD1, luciferase and EGFP genes with partial

5’- and 3’-UTRs, obtained from RT-PCR product of hSOD1, pGL3 vector, and

pEGFP-N2 vector respectively, were subcloned into the pcDNA3 vector using the

following sets of restriction enzymes: BamHI and NotI for hSOD1 and EGFP, HindIII

and XbaI for luciferase.

4.3.3 In vitro transcription and oxidation

RNAs, which were synthesized in vitro, were considered to be non-oxidized. PCR

was performed to pcDNA3-luciferase, pcDNA3-EGFP or pcDNA3-hSOD1 plasmid

by using T7+ and SP6-(dT)30 primers (T7+: 5’-

GGATCCTAATACGACTCACTATAGG; SP6-(dT)30: 5’-

(dT)30CATTTAGGTGACACTATAG). After purification, the amplified cDNAs were transcribed by T7 RNA polymerase (US Biologicals) to RNAs containing poly(A) tail and m7G(5’)ppp(5’)G RNA capping analog (Invitrogen). For oxidation, the capped

poly(A)-tailed RNAs were incubated with H2O2 and cytochrome c (Sigma, St. Louis,

MO) at different dosage at 37 oC for 1 hour. RNA was purified by RNeasy mini kit

(Qiagen).

4.3.4 In vitro translation

85

In vitro translation using Rabbit Reticulocyte Lysate system (Promega, Madison, WI)

was performed according to the manufacturer’s direction on in vitro generated

luciferase RNAs, both non-oxidized and oxidized, for 1.5 hours at 30 oC. Luciferase activity was examined using Luciferase Assay System (Promega). Biotinylated luciferase was created by incorporating biotinylated lysine residues into nascent proteins during translation using the TranscendTM Non-Radioactive Translation

Detection Systems (Promega). Horseradish peroxidase avidin D (Vector Laboratories,

Burlingame, CA) was used to visualize the biotinylated proteins followed by

chemiluminescent detection.

4.3.5 RNA transfection

Cells were seeded to 80-90% confluence in 6-well tissue culture plates. 2 µg of in

vitro synthesized and/or oxidized mRNAs were diluted in serum-free OPTI-MEM

(Invitrogen) containing 6 µl DMRIE-c reagent (Invitrogen). After incubating with

RNA mixture for 6 hours at 37 oC, cells were cultured in normal complete cell media for another 16 hours. EGFP protein expression was then observed under fluorescence

microscope. Luciferase or hSOD1 mRNA-transfected cells were harvested for

functional assay and/or SDS-PAGE.

4.3.6 Plasmid DNA transient transfection

HEK cells were seeded to 50% confluence in 6-well plates and were transiently transfected with pcDNA3-hSOD1 vectors using LipofectAMINE PLUS (Invitrogen)

86

according to standard protocols. We used 6 µl PLUS and 4 µl lipofectAMINE for

each well. Cells were harvested 48 hours after transfection.

4.3.7 Total RNA isolation by TRIZOL

Total RNA was isolated from transfected cells using TRIZOL (Invitrogen) according

to manufacturer’s directions. Briefly, 1 ml of TRIZOL per well in 6-well plates was

added to cultures, and incubated at room temperature for 5 min. 0.2 volumes of

chloroform was added to the sample and mixed by vigorous shaking for 15 sec and

incubation at room temperature for 3 min. The organic phase was separated from the

aqueous phase by centrifugation at 12,000xg for 15 min at 4 oC. The aqueous phase

was transferred to a new tube and the RNA was precipitated with 2-propanol (0.5

volumes; 10 min at room temperature). The RNA was pelleted by centrifugation at

12,000xg for 10 min at 4 oC. The RNA pellet was then washed with 1 volume 70%

ethanol and pelleted by centrifugation at 7,500xg for 5 min at 4 oC. The RNA pellet

was air dried and then resuspended in 20-100 µl DEPC-treated ddH2O.

4.3.8 Quantitative RT-PCR

To examine the stability of transfected RNAs, we collected RNAs at different time

points (4, 8, 16, 24, 32, 48 and 72 hours) after RNA transfection. Total RNAs were

isolated from cells directly by Trizol (Invitrogen). Reverse transcription (RT) was

performed by using (dT)30 primer and Superscript II reverse transcriptase (Invitrogen)

following the manufacturer’s protocol. PCR was then performed on the RT products

87

using specific primers for β-actin and EGFP F (5’-atggtgagcaagggcgaggag-3’), R (5’-

gtccatgccgagagtgatccc-3’) with different amplification cycle numbers (20, 25, 30

cycles).

4.3.9 Luciferase assay

Cells were rinsed by PBS and lysed in 1X luciferase lysis buffer (Promega). The

lysates were assayed for luciferase activity using GENios microplate reader (Tecan

Co., San Jose, CA).

4.3.10 Immunoprecipitation

Cells were harvested 16 hours after RNA transfection. Cells were solubilzed in PBS

containing Complete protease inhibitor cocktail (Roche Applied Science) with 1%

Triton X-100 at 4 oC for 30 min with gentle shaking. Cell debris was removed

(15,000g, 15 min) and supernatants were incubated with protein G agarose beads and

rabbit anti-luciferase polyclonal antibody (1:100; Cortex Biochem, San Leandro, CA)

at 4 oC for 16 hours. The immunobead-bound protein complexes were washed three times by PBS buffer plus 0.1 mM PMSF and 5 mM EDTA in 1% Triton X-100 followed by washing with PBS buffer plus 0.1 mM PMSF and 5 mM EDTA for

another three times.

Biotin-labeled luciferase proteins in rabbit reticulocyte lysates were diluted in 1x PBS

containing Complete protease inhibitor cocktail (Roche Applied Science) with 1%

Triton X-100 and were incubated with Immobilized Neutravidin Binding Protein

(Pierce) at 4 oC for 16 hours. The immunobead-bound protein complexes were

88

washed three times by PBS buffer plus 0.1% Triton X-100 and were resolved in SDS-

PAGE.

4.3.11 Western Blotting

Proteins were separated on a 7.5% polyacrylamide gel containing 0.1% SDS by electrophoresis and then electrotransferred onto a polyvinylidene difluoride (PVDF) membrane (0.45 µm, Roche Applied Science). The membrane was incubated with

HRP-conjugated Avidin D in PBS plus 0.5% Tween-20 for 60 min at room temperature or rabbit polyclonal antibody against luciferase or hSOD1 at 1:1000 dilution in 1% nonfat milk in PBS plus 0.1% Tween-20 overnight at 4°C followed by peroxidase-conjugated goat anti-rabbit IgG (1:1200; ICN Biomedicals, Aurora, OH) for 60 minutes at room temperature. The immunoreactive bands were detected using the LumiLight Western Blotting Substrate (Roche Applied Science) according to manufacturer’s directions.

4.3.12 Polyribosome association

Sucrose gradients (15 to 60%) were prepared by layering 1.5 ml of 15% sucrose on top of an equal volume of 60% sucrose in Beckman polyallomer centrifuge tubes

(Beckman). Gradients ere formed by placing tubes in a horizontal position for 5 hr

(Davies and Abe, 1995). 50 ml of translational productes in rabbit reticulocyte lysates diluted in 300 ml of buffer (100 mM KCl, 2 mM DTT, 2 mM magnesium acetate, 20 mM Tris-HCl, pH 7.5) were layered onto a sucrose gradient and spun in a Beckman

89

SW40 rotor at 30,000 rpm for 3 hr at 4 °C. Ten fractions were collected per tube,

starting from the top, and were mixed thoroughly with equal volume of phenol-

chloroform-isoamyl alcohol. The resulting mixtures were centrifuged at 14,000rpm

for 5 min and the aqueous phases after centrifugation were precipitated with 5 µg of glycogen, 2 volumes of 95% ethanol and 0.1 volume of 3M acetic acid (pH 5.2).

Precipitaed RNAs were resuspended in 5 µl of DEPC-treated ddH2O for Northern blotting analysis.

4.3.13 Northern Blotting

DIG-labeled cRNA probes were synthesized by in vitro transcription using 1 µg of

DNA template (linearized pcDNA3-18S rRNA sequence), DIG-11-UTP (350 µM;

Roche Applied Science) and SP6 RNA polymerase (100 U; U.S. Biochemicals,

Cleveland, OH). In vitro transcription proceeded for 2.5 h at 37°C after which the template was digested with (5 U; Invitrogen) for 15 min at 37 °C.

DIG-labeled DNA probes were generated from luciferase PCR products using

Digoxogenin High Prime DNA Labeling and Detection Starter Kit II (Roche Applied

Science). mRNAs were fractionated by electrophoresis through a 1% agarose/2.2 M formaldehyde gel and then transferred to positively charged nylon membranes by over-night capillary transfer with 10x SSC. The RNA was then fixed to the membrane by crosslinking as above followed by baking for 30 min at 80°C. The membranes were incubated with modified Church buffer (10% SDS, 1 mM EDTA,

2% blocking reagent (Roche Applied Science) in 250 mM Na2HPO4, pH 7.2

90

(Shifman and Stein, 1995)) for 3 h at 68°C and then with DIG-labeled cRNA probe or

DNA probe in modified Church buffer overnight at 68°C or 55°C respectively. The

membranes were subsequently washed three times for 20 min each wash at 68°C with

washing buffer (1% SDS and 1 mM EDTA in 25 mM Na2HPO4, pH 7.2). Detection

of DIG-labeled probe hybridization was accomplished using Digoxogenin High

Prime DNA Labeling and Detection Starter Kit II according to the manufacturer’s

direction

4.3.14 mRNA microinjection

mRNAs were in vitro synthesized from EGFP DNAs or isolated from HT4 cells, and

were then in vitro oxidized with 3.2 µM H2O2 and 0.2 µM cytochrome c. After purification, mRNAs was quantified at 260 nm UV light using spectrophotometer

(Beckman) and were adjusted to serial concentrations including 0.25, 0.5, 1, and 2 µg/

µl. Fluorescent dyes were included in mRNA samples prepared from HT4 cells.

mRNA microinjection was accomplished with Eppendorf microinjector. Each

injection delivered 0.5 pl of solution into cytoplasm of HT4 cells. Injected cells were

observed under fluorescent microscope 24 hr after injection by monitoring

fluorescence from EGFP proteins or dyes.

4.4 Results

4.4.1 RNA oxidation causes decreases of expression level and activity of the protein

in rabbit reticulocyte lysate.

91

First, we applied gene transcripts of luciferase and enhanced green fluorescence

protein (EGFP) to study the consequence of oxidized mRNAs. These two species

were selected because of good resources to examine the expression levels and

activities of the proteins. We synthesized luciferase mRNA in vitro and then oxidized

the mRNAs to various degrees using different amounts of H2O2 and cytochrome c.

More than 80% of the mRNA was oxidized when treated with 1.6 µM H2O2 and 0.1

µM cytochrome c as determined by immunoprecipitation. We used a rabbit

reticulocyte lysate to translate the oxidized luciferase mRNAs into proteins. The expression levels and activities of the protein were examined by analysis and functional assay, respectively. As shown in Figure 4.1, there was a dramatic, dose-dependent loss of the detectable protein levels (A) as well as luciferase activities

(B) in the oxidized mRNA samples. We measured the signal intensities of each band in Figure 4.1A by densitometry. The data showed a positive correlation between the decrease of protein levels in oxidized mRNA samples and the decrease of the luciferase activity.

92

Figure 4.1 Oxidation of luciferase mRNA decreases protein expression and functional activity. Firefly luciferase mRNAs oxidized with serial concentrations of oxidants, H2O2 and cytochrome c, were expressed in rabbit reticulocyte lysates. The data showed that detectable luciferase protein levels (A) as well as luciferase activities (B) decrease in the oxidized mRNA samples in a dose-dependent manner. The protein levels were determined by visualizing luciferase proteins labeled with biotinylated lysine residues by Western blot analysis. The protein activity was determined by luciferase function assay.

4.4.2 RNA oxidation causes decreases of expression level and activity of the protein in cell lines.

To further confirm the above observation, we expressed the oxidized luciferase mRNAs in HEK293 cells by transfection. The transfected cells were harvested for functional assay and protein expression detection 16 hours after transfection.

93

Consistent with the rabbit reticulocyte lysate results, we observed a dose-dependent loss of luciferase activities and the protein levels in the oxidized mRNA samples

(Figure 4.2). The similar results were also observed in PC12 cells.

Figure 4.2 Expression of oxidized luciferase RNA in HEK cells. (A) A dose- dependent loss of luciferase activities in the oxidized mRNA samples was also observed in cell lysate from luciferase RNA-transfected HEK cells. (B) The protein levels were determined by immunoprecipitation using anti-luciferase antibodies followed by immunoblotting. The arrow indicates expressed luciferase proteins; the last lane is purified luciferase protein.

94

To detect protein expression directly in living cells, we used enhanced green fluorescent protein (EGFP) to study the consequence of mRNA oxidation. EGFP mRNAs were synthesized in vitro and oxidized with 3.2 µM H2O2 and 0.2 µM cytochrome C. Both non-oxidized and oxidized EGFP mRNAs were transfected into

HEK293, a human embryonic kidney-derived cell line, as well as Neuro2A, a mouse neuroblastoma cell line. A decrease of EGFP expression was also observed in oxidized mRNA-transfected cells 16 hours after transfection under fluorescence microscope (Figure 4.3).

Figure 4.3 Expression of EGFP mRNA in HEK293 and Neuro2A cell lines. HEK293 and Neuro2Acells were transfected with non-oxidized or oxidized in vitro synthesized EGFP mRNAs. The decrease of EGFP protein expression, indicated by weaker fluorescence, was detected in oxidized mRNA-transfected cell.

95

The studies on the consequence of luciferase and EGFP mRNA oxidation showed

the decreases of protein level and activity. Next we investigated whether oxidized

mRNA species found in AD brains had the same consequence. Human Cu/Zn SOD 1

(hSOD1) mRNA is a highly oxidized species in AD as shown in Table 2 with 66.6% of its transcripts being oxidatively modified. We transfected both non-oxidized and oxidized hSOD1 mRNAs in HEK293 cells and harvested cells 16 hours after transfection. mRNA was oxidized with 3.2 µM H2O2 and 0.2 µM cytochrome C. The

protein level was examined by Western blotting analysis (Figure 4.4). We consistently observed a decrease of hSOD1 protein expression in cells transfected with oxidized hSOD1 mRNA. The level of β-actin did not show difference, indicating the equal loading of samples.

Figure 4.4 Oxidation of hSOD1 mRNA results in a decrease of detectable protein level of hSOD1 in HEK293 cells. HEK293 cells were transfected with pcDNA3 plasmid carrying hSOD1 gene (control), non-oxidized (non-ox) and oxidized (ox) hSOD1 mRNA. Omission of mRNA during transfection (no RNA) was considered as a negative control, in which only endogenous hSOD1 protein was present. β-actin was used as the internal control. 20 µg of protein lysates from RNA- transfected cells and 5 µg of protein lysates from plasmid DNA-transfected cells were loaded on SDS-PAGE.

96

To investigate whether the decline of protein function and Western-blot

detectable protein level is the result of rapid degradation of the oxidized mRNAs after

entering the cells, we examined the levels of oxidized EGFP and hSOD1 mRNAs at different time points after transfection using quantitative RT-PCR. The oxidized mRNA levels remained similar to the non-oxidized mRNA levels for as long as 72 hours (Figure. 4.5). The transfected hSOD1 mRNAs was distinguished from

endogenous hSOD1 mRNA by incorporating SP6 sequence from pcDNA3 vector at

3’-untranlational region (UTR) of hSOD1 gene constructs. The result was further

confirmed by Northorn blot analysis of transfected-hSOD1 mRNAs. These results

indicate that the oxidized mRNAs may mediate the decrease of protein expression by

impairing the normal process of protein translation.

97

Figure 4.5 Detection of stability of oxidized and non-oxidized mRNAs at different time points after transfection by quantitative RT-PCR analysis. HEK293 cells were transfected with either oxidized mRNAs (O; treated with 3.2 µM H2O2 and 0.2 µM cytochrome c) or non-oxidized mRNAs (N) of EGFP (A) or hSOD1 (B). Cells were harvested at indicated time points and total RNAs were subsequently isolated for quantitative RT-PCR analysis. β-actin mRNA was used as an internal control. Negative controls were included: omission of mRNAs (no RNA) during transfection and omission of transfection reagent DMRIE-c during transfection (no transfection). The oxidized mRNAs remained at similar levels as non- oxidized mRNAs for up to 48 hr examined in EGFP mRNA-transfected cells and 72 hr in hSOD1 mRNA-transfected cells.

4.4.3 RNA oxidation causes the abnormal increase of polyribosome association

RNA oxidation causes a decrease of normal protein expression. This does not result from the rapid degradation of oxidized mRNAs in cells as proved by mRNA stability experiment (Figure 4.5). Then, how the oxidative modifications on mRNA transcripts affect the downstream translational process is the next issue to study. We compared polyribosome associations on non-oxidized and oxidized mRNA transcripts.

Luciferase mRNAs were in vitro oxidized with 3.2 µM H2O2 and 0.2 µM cytochrome

98

c. Both non-oxidized and oxidized luciferase mRNAs were incubated in rabbit

reticulocyte lysate with other essential components to process translation reaction for

10 min. The lysates were loaded and fractionated in 15-60% sucrose gradient by

centrifugation. Ten fractions were collected per tube from the top. The

monoribosomes remained on the top of the gradient in the region of less than 20%

sucrose, while polyribosomes were preferentially found at sucrose concentration

>35%, as measured by UV tracing at 254 nm and sucrose refractometry (Dickey et al.,

1998). mRNA transcripts associated with more polyribosomes located at heavier

fractions of sucrose gradients. After extraction of RNA from individual fractions,

Northern blotting was performed to quantify interested RNA species using specific anti-sense RNA probes or DNA probes. As shown in Figure 4.6A, lucferase mRNAs,

both non-oxidized and oxidized, was examined in different fractions. Compared to

non-oxidized luciferase mRNAs oxidized ones had stronger intensity of luciferase

mRNA bands at heavier fractions. 28S and 18S rRNAs were also examined on the

blots. Our data indicated that fraction 4 primarily contains monoribosomes and fractions 5-10 have polyribosomes. mRNA oxidation correlated with an increased polyribosome association of the mRNA messages. We measured signal intensities for luciferase mRNAs in each fraction by densitometry. The abundance of luciferase mRNA was expressed by calculating the proportion of luciferase mRNAs from each fraction to the total amount of luciferase mRNAs. As shown in Figure 4.6B, the peak of signal intensity was located at fraction 8 in non-oxidized luciferase mRNA- containing lysate, while the peak was shifted to fraction 9 in oxidized luciferase-

99

containing lysate. We repeated this experiment three times and consistently observed

the peak shift to heavy fraction in oxidized mRNA trials. We performed the same

experiment to mRNAs oxidized with higher concentrations of oxidants (8 µM H2O2 and 0.5 µM cytochrome c). The result showed that more oxidized luciferase mRNAs moved to the bottom fraction 10, indicating that more polyribosomes associated on these oxidized transcripts. The data from polyribosome association expression suggests that oxidized bases on mRNAs may cause ribosome stalling on the transcripts.

100

Figure 4.6 Polyribosome profiles of non-oxidized and oxidized luciferase mRNAs in rabbit reticulocyte lysates. (A) Non-oxidized and oxidized luciferase mRNAs were processed with protein translation in lysates which were then fractionated by 15-60% sucrose gradient. Ten fractions were collected for each sample and RNAs from each fraction were resolved by gel electrophoresis, blotted, and probed with DIG-labeled antisense RNA or cDNA to specific regions of interested genes. (B) The signal of luciferase mRNA in each fraction shown in (A) was quantified by densitometry and those for the whole gradient were set as 100%. The value for each fraction was expressed as percentage of the total gradient.

101

4.4.4 No truncated protein was generatied from oxidized mRNA examined by

immunoprecipitaion and Western blotting.

To investigate whether oxidized sites on mRNA transcripts cause premature

termination of translation and thus generation of truncated proteins, we studied the

translation products of non-oxidized and oxidized luciferase mRNAs using polyclonal

luciferase antibody against the whole purified luciferase protein. We performed

Western blotting analysis using SDS-PAGE with different concentrations of

polyacrylamide. No extra band with smaller size other than the one indicating full-

length luciferase protein appeared (Figure 4.7). Two batches of polyclonal luciferase

antibodies from different vendors were used and consistent results were obtained. To

avoid the possibility that luciferase antibodies missed the detection of truncated

proteins, biotin-labeled lysines were incorporated into luciferase proteins during the

translation and luciferase proteins were detected with HRP-conjugated Avidin. As shown in Figure 4.7, no truncated product was identified by immunoreactivity of

HRP-conjugated Avidin. We considered that truncated proteins, if produced during the translation of oxidized luciferase mRNA, may be only present in certain amount that was too low to be detected by the Western blotting analysis. We applied

immobilized NeutrAvidin to pull down luciderase proteins from lysates in which biotin-

labeled proteins were translated from non-oxidized and oxidized luciferase mRNAs.

Precipitated proteins were examined by Western blotting with luciferase antibody

(Figure 4.7, upper panel) and avidin (Figure 4.7, lower panel). Nonspecific immunoreactivity with luciferase antibody and avidin D was observed in every lysate

102

sample but this was not detected in the precipitated samples. No truncated luciferase protein was detected in these experiments, indicating that oxidized mRNA did not cause early termination of translation. Alternatively, this technique may not be able to detect the truncated luciferase protein in a small quantity.

Figure 4.7 Western blotting to examine the presence of truncated proteins translated from oxidized luciferase mRNA. Luciferase proteins were translated from either non- oxidized (non-ox) or oxidized mRNAs (ox; treated with 3.2 µM H2O2 and 0.2 µM cytochrome c) in rabbit reticulocyte lysate with biotin-labeled lysine. The proteins in lysates were detected by SDS-PAGE using either anti-luciferase antibody or HRP-conjugated Avidin D. Luciferase proteins was immunoprecipitated with anti-luciferase antibody (IP; upper panel) or immobilized NeutrAvidin D (IP; lower panel) and then visualized by SDS-PAGE. Omission of mRNA (no RNA) during the translation was used as the negative control, and translation products from luciferase mRNA provided in the lysate kit were used as the positive control (control). No distinguishable truncated protein was detected in oxidized mRNA-translated products.

103

4.4.5 Investigation of protein mutation from oxidized mRNA by SELTI-TOF.

Oxidized nucleotides in DNA can cause single-base mutation during replication.

We considered whether oxidized bases in mRNA could lead to amino acid

substitution by mispairing between oxidized mRNA and tRNA during translation. To

study this possibility we performed surface-enhanced laser desorption and ionization- time of flight mass spectrometry (SELDI-TOF), which can distinguish minor

differences of amino acid constitutions of targeted proteins by mass change. We in

vitro oxidized luciferase mRNA by 3.2 µM H2O2 and 0.2 µM cytochrome c and

expressed non-oxidized and oxidized mRNAs in rabbit reticulocyte lysates.

Translated luciferase protein was immunoprecipitated by polyclonal anti-luciferase

antibody and protein G. acetic acid (1 M) in 50% acetonitrite was used to elute

luciferase proteins off from the antibody and protein G complex. This elution

condition can also wash off some antibody proteins into eluates. One control trial in

which no luciferase mRNA was included in the translational reaction and the lysate

was immunoprecipitated with anti-luciferase antibody was performed in the same

manner. The eluates were sent to Proteomics facility at Heart and Lung Research

Institute for SELDI-TOF analysis. Two sets of protein chips with different matrixes

were used: one is cyano-4-hydroxy cinnamic acid (CHCA) and the other is sinapic

acid (SPA). Unfortunately, only four peaks which represent the complex of one heavy

chain and one light chain of IgG with the mass of 72 kD, the heavy chain at 48 kD,

the fragmentation products of heavy chain at 36 kD and 12 kD were identified in the

104

CHCA matrix chips. No peak indicating luciferase protein with 61 kD or the light

chain of IgG with 24 kD were detected by SELDI-TOF.

4.4.6 mRNA oxidation could be lethal to cells.

The magnitude of oxidized mRNAs can reach 30-70% in AD brains, and neuron

is the major cell type bearing mRNA oxidation. Therefore, we can estimate that the level of oxidized mRNA in neurons is even higher than that measured from trunks of

AD brain tissues. This initiated the following question that is how cells react with high concentration of oxidized mRNA. In our previous study, we transfected oxidized mRNA into cells, but the amount of transfected mRNAs was not very high because of the limited transfection efficiency. Then we applied RNA microinjection technique to

investigate the effect of oxidized mRNAs at different concentrations by collaboration

with Dr. Chanden Sen’ lab (Heart and Lung Research Institute, the Ohio State

Univeristy). EGFP mRNAs were in vitro oxidized with 3.2 µM H2O2 and 0.2 µM

cytochrome c. Non-oxidized and oxidized EGFP mRNAs, quantified and then diluted into serial concentrations of 0.1, 0.25, 0.5, and 1 µg/µl, were injected into HT4 cells.

HT4 is a human hippocampal neuron-derived cell line. Each injection delivers 0.5 pl of solution into the cytoplasm of cells. 24 hour after microinjection, cells were observed under fluorescent microscope. Interestingly, non-oxidized EGFP mRNAs, up to the concentration of 1 µg/µl, were nicely expressed in heathy-looking HT4 cells,

while oxidized EGFP mRNAs, at above the concentration of 0.5 µg/µl, led to cell

death 24 hours after injected into HT4 cells (Figure 4.8). 20 cells were injected with

105

non-oxidized mRNAs or oxidized mRNAs in one experiment. This was repeated three times with similar results.

non-oxidized EGFP mRNA oxidized EGFP mRNA

0.1 µg/µl

0.25 µg/µl

0.5 µg/µl

1 µg/µl

Figure 4.8 Delivery of oxidized EGFP mRNAs into HT4 cells causes cell death. Non- oxidized and oxidized EGFP mRNAs at different concentrations were microinjected to HT4 cells. Cells were observed under fluorescence microscope 24 hr after injection. Oxidized EGFP mRNAs at 0.5 µg/µl and higher caused cell death, while non- oxidized EGFP mRNAs did not change cell viability.

106

To investigate whether oxidation to endogenous mRNAs had the same effect as oxidized EGFP mRNAs, we isolated mRNAs from HT4 cells and in vitro oxidized them with 3.2 µM H2O2 and 0.2 µM cytochrome c. Both non-oxidized and oxidized

mRNAs from HT4 cells were mixed with non-toxic fluorescent dye to facilitate the

localization of injected cells. mRNA samples were diluted in serial concentrations

and then were microinjected into HT4 cells. Again oxidized HT4 mRNAs caused cell

death at the concentration of 0.25 µg/µl and up, while non-oxidized HT4 mRNAs did

not (Figure 4.9).

107

non-oxidized HT4 cell mRNA oxidized HT4 cell mRNA

0.1 µg/µl

0.25 µg/µl

0.5 µg/µl

Figure 4.9 Delivery of oxidized mRNAs of HT4 cells causes cell death. mRNAs were isolated from HT4 cells and oxidized in vitro. Both non-oxidized and oxidized mRNAs at different concentrations were microinjected to HT4 cells with fluorescent dyes. Cells were observed under fluorescence microscope 24 hr after injection. Oxidized mRNAs at 0.25 µg/µl and higher caused cell death, while non-oxidized mRNAs did not change cell viability.

4.5 Discussion

To my knowledge, this is the first study that investigates the consequence of

oxidized mRNAs. We used luciferase and EGFP mRNAs in the initial study because

of commercial availability of good antibodies, functional assays and the convenience

108

of detection. We observed that (1) oxidized luciferase and EGFP mRNAs are as

stable as the non-oxidized mRNA in the HEK293 cells, (2) oxidation of luciferase

and EGFP mRNAs leads to loss of normal protein level and protein function in the rabbit reticulocyte lysate as well as in the HEK293 and Neuro2A cells, (3) luciferase mRNA oxidation causes an increased association of polyribosomes on messages, and

(4) oxidation to EGFP mRNAs or endogenous mRNAs of HT4 cells can be lethal to

HT4 cells.

Cells may protect mRNA from oxidation via three pathways. First, intracellular antioxidant system, in which Cu/Zn SOD1, catalase and glutathione peroxidase are the major enzymes, modulates the production of ROS. Studies showed that antioxidant enzymes were dysregulated in AD brains. Neurons in AD affected regions were under the attack of increased ROS. mRNA can be the major target because of its widespread subcellular distribution, relative abundance, single strand nature and lack

of protection from histone proteins. Second, the discovery of RNA repair enzymes

rectified the conventional thought that RNA can not be repaired. mRNA repair

mechanisms have long been underestimated because mRNAs are present in cells in

several to thousands copies of individual species and are replaceable. Until recently,

an RNA repair enzyme hABH3, which repairs the methylated mRNAs, has been

reported (Aas, et al., 2003). Although no repair enzyme for oxidized mRNA has been

studied yet, there is a possibility that such kind of enzyme is present in cells since

mRNA oxidation occurs not less frequently than its methylation. The alteration of

109

such kind of enzyme, if existing, would be studied in AD brains to better understand mRNA oxidation in AD. Finally, oxidized mRNAs may be recognized by RNA degradation system therefore removed from the cytoplasm. In the study of the consequence of oxidized mRNAs, we found that mRNA oxidation leads to a decrease of normal protein expression. Is this the result of degradation of oxidized mRNAs in cells? We studied the stabilities of non-oxidized and oxidized mRNAs introduced into cells by transfection. The result showed that oxidized mRNAs are as stable as non- oxidized ones. This indicates that the decrease of normal protein expression from oxidized mRNAs can result from abnormal processesing in protein translation.

Another possibility is that the availability of oxidatively modified mRNAs to translational processes is decreased by binding transcripts with RNA-binding proteins, such as Y-box-binding protein 1 (YB-1), which is a nucleic-acid-binding protein that forms a major structural component of messenger ribonucleoprotein particles and involves in regulating protein translation (Kohno et al., 2003). Binding with YB-1 protein, 8-OHdG-containing mRNA can be removed from the pool of translational materials to diminish aberrant protein production (Hayakawa et al., 2002; Ougland et al., 2004). However, our study showed that the low translational activity of oxidized mRNA message is associated with an increase of polyribosome loading on transcripts.

This suggests that mRNA oxidation has deleterious effect on protein translational process.

110

Protein translation consists of three steps: initiation, elongation and termination.

Oxidized mRNAs may impair the translation process in any of three steps depending on the localization of oxidized bases in the transcripts. Oxidation on ribosome small subunit-binding site and/or its downstream may change the affinity of ribosomes with mRNA transcripts and the migration rate of ribosome small subunits in 5’- untranslated region (UTR) of mRNAs. mRNAs with oxidative lesions in coding regions may have a reduced efficiency for the oxidized bases in the transcripts to form base pairs required for proper recognition by tRNA anticodon. Also the elongation rate may change if ribosome complexs can not pass through the oxidized sites of mRNAs efficiently. Oxidation of bases encoding stop codons in mRNAs may interfere with normal termination procedure. Oxidation in 5’- or 3’- UTRs of mRNAs may affect the translational regulation controlled by these regions. It is likely that the presence of oxidation lesions in coding regions or stop codons of mRNAs contributes to ribosome stalling observed by polyribosome association experiment (Figure 4.6).

Some studies showed that ribosome stalling can occur to truncated mRNAs lacking stop codons. Both prokaryotic and eukaryotic cells have elaborate mechanisms to degrade the translational products of such transcripts, thereby liberating stalled ribosomes and preventing the generation of potentially harmful, truncated proteins

(Karzai et al., 2000; Keiler et al., 1996). However, this kind of fine regulation pathway may not be able to accommodate a great amount of damaged mRNAs as 30-

70% of mRNAs are oxidatively modified in AD brains (Shan et al., 2003). Oxidized mRNAs may mediate their deleterious effects by stalling ribosomes on their

111

transcripts, consuming the intracellular ribosome pool in addition to decreasing the

production of normal corresponding proteins. Are truncated proteins released from

the improper translation of oxidized mRNA? We investigated this issue by Western

blotting in combination of immunoprecipitation. We did not detect specific signals of

interested proteins with smaller sizes. It is possible that oxidized mRNAs did generate

truncated proteins with different sizes which are present in a very small quantity

individually and our detection method is not sensitive enough to visualize them. We

also considered whether oxidized mRNA generates mutant proteins. We compared

protein levels and activities of luciferase translated from both non-oxidized and

oxidized luciferase RNAs. The decrease of protein activity is correlated to the

decrease of protein expression based on the quantitative analysis. However, we can

not rule out the possibility of generation of mutant proteins since the majority of

oxidized sites on luciferase mRNAs may be outside of the enzymatically functional

site of the luciferase protein. Therefore we applied SELDI-TOF to examine whether

mutant proteins were produced from oxidized luciferase mRNAs. Theoratically, the

protein containing mutants can generate multiple peaks in addition to the one with

normal protein mass, or one broader peak compared to the normal protein on the

spectrum of SELDI-TOF. There are three possibilities to explain the failure of

detection of luciferase proteins by SELDI-TOF. First and most likely, the quantity of

luciferase proteins isolated from rabbit reticulocyte lysates was too limited. Based on the report from the manufactory, about 20 ng of proteins can be generated in 50 µl of lysate. However, we cannot accurately determine the protein level in our. We purified

112

the luciferase proteins after translation of luciferase mRNA in the lysate and

examined the protein by sliver staining of SDS-PAGE, which can visualize proteins

above 4-10 ng in the gel. We did not detect any positive signal for luciferase proteins.

The second possibility is that the luciferase proteins can not bind well on the cyano-4- hydroxy cinnamic acid-coated suface of protein chips that we used for SELDI-TOF.

Finally, the luciferase protein may not be ionizied well, leading to the inability of detection by mass spectrometer. The experiment may be accomplished by increasing the volume of lysates used for translation to elevate the yield of interested proteins and/or applying other chemical- or antibody-coated protein chips to run SELDI-TOF.

Another method applicable to examine the sequence of the protein is Tandem Mass

Spectrometry (MS-MS), which can give the detailed information about truncated proteins and the amino acid substitutions of mutant proteins.

We demonstrated here that delivery of oxidized mRNAs into HT4 cells by microinjection causes cell death. Cells can survive from the low intracellular concentration of oxidized mRNAs. Once the amount of oxidized mRNAs reaches a certain level (0.125 pg/cell), cell death occurs. The neurotoxic effects may be mediated by oxidized mRNAs themselves and/or their aberrant protein products.

Studies have shown that RNA alone can be pathogenic to alter cellular functions in several human diseases (Ranum and Day, 2002). Oxidized mRNAs may mediate a toxic gain-of-function by attracting and sequestering ribosomes, RNA-binding proteins, RNA degradation system or RNA repair enzymes from their normal

113

functions. It is also possible that different oxidized mRNA species may behave differently and have different effects depending on their natures. Further investigation should be performed to better understand the involvement of mRNA oxidation in

neurodegeneration. Compared to DNAs, oxidative damages to mRNAs may have a

limited impact on cell function because of its relative abundance, its expendability

and inability to be inherited. However, significant oxidation of mRNAs in AD

appears to be a potential contributor to the pathogenesis of the disease.

114

CHAPTER 5

INVESTIGATION OF RNA OXIDATION IN CELL MODELS

5.1 Abstract

We previously discovered that mRNAs are highly oxidized in brains of

Alzheimer’s disease (AD) and mRNA oxidation causes abnormal protein translation.

In this study, we investigated RNA oxidation in primary cortical neuronal cultures to

study the relationship of RNA oxidation and neuronal cell death. Our results showed

that RNA oxidation is an early event far preceding cell death induced by different insults, including H2O2, glutamate, and amyloid β peptide, which are implicated in

the pathogenesis of AD. Neurons, not astrocytes, are specifically susceptible to RNA

oxidation under oxidative stress. Some mRNA species highly oxidized in AD brains

such as human Cu/Zn SOD1 and glutamate dehydrogenase 1 are also present in

oxidized mRNA pool of oxidative insult-treated cultures. The protein expression level

of oxidized mRNA species of human Cu/Zn SOD1 was decreased in cultures. Our

data suggests that RNA oxidation may have functional importance in the cascade of

neuronal cell death and occurs through the impairment of protein translation.

115

5.2 Introduction

An increasing body of evidence implicates the involvement of oxidative stress in

Alzheimer’s disease (AD) (Markesbery, 1997; Lovell et al., 1997, 1999, 2000; Pratico

and Sung et al., 2004). Oxidative stress is defined as an imbalance of the oxidant- antioxidant ratio in favor of the former, leading to cell damages (Sies, 1991). The central nervous system is highly vulnerable to oxidative imbalance on account of its high rate of oxygen consumption, enrichment of polyunsaturated fatty acids, relative paucity of antioxidant system, and high content of transition redox metals. Depending on the substrates attacked by reactive oxygen species (ROS), oxidative damages in

AD brains have been manifested as lipid peroxidation, protein oxidation, DNA oxidation and RNA oxidation. Traditionally, oxidative damages to intracellular molecules are viewed as the result of neurodegeneration. However, some studies have pointed out that oxidative damages may be functionally important in the pathogenesis of AD. Studies on oxidation of proteins and RNAs in AD demonstrated that the increased oxidative modification occurs to some morphologically intact neurons in

AD and is not consistently associated with mature senile plaques, suggesting that oxidation of intracellular macromolecules precedes neurodegenerative changes

(Aksenov et al. 2001; Nunomura et al., 1999). Pratico et al. (2004) investigated a marker of lipid peroxidation in vivo using the samples of urine, plasma, and cerebrospinal fluid (CSF) from living patients in combination with clinical diagnosis of AD. They found that oxidative damage to lipids is an early event during the development of the disease. They also studied the subjects with mild cognitive

116

impairment (MCI), which is associated with up to a 50% probability of progressing to

symptomatic AD within a 4-year period (Petersen et al., 2001). The result showed

that, similar to AD subjects, patients with MCI also have the increased level of the

marker of lipid oxidation in samples of urine, plasma and CSF as compared to

cognitively normal elderly controls. However, no significant difference of the

contents of tau and Aβ42 in CSF, which are common markers of AD neuropathology

and disease progression (Andreasen et al., 2001) and are elevated and decreased, respectively, in AD cases, was observed between MCI subjects and controls. This

observation demonstrated that oxidative damages are present before the detection of

the typical neuropathological changes of early AD. A conclusion from these

observations is that the ROS-induced damage to intracellular molecules is not a

simple consequence of cell death but may play an important role in AD pathogenesis.

RNA oxidation is a prominent feature in vulnerable neurons in AD affected

regions (Nunomura et al., 1999). Our previous data showed that 30-70% of mRNAs

are oxidized in AD frontal cortices (Shan et al., 2003). In this study, we investigated

the relationship of RNA oxidation and cell death by the measurement of time courses

of RNA oxidation in neuronal cultures under oxidative stress, which is implicated in

AD pathogenesis. We found that RNA oxidation is an early event far preceding cell death. Neuron is the major targeted cell type of oxidative damages although the resource of elevated ROS generation, which is neuron, astrocyte, or microglia, remains controversial (Almeida et al, 2002; Kress et al., 2002; Abramov et al., 2004;

117

Meda et al. 1995). Our data showed that RNA oxidation is primarily localized in some group of neurons but not in astrocytes. The study on the effect of oxidized mRNA species in the culture demonstrated the decrease of protein expression.

Oxidized mRNAs may mediate this effect and impair normal functions in neurons.

5.3 Materials and Methods

5.3.1 Reagents

30% H2O2 stock solution was purchased from Fisher (Fair Lawn, NJ). Aβ25-35

was purchased from Bachem (Torrance, CA ). Papain, glutamate, poly-D-lysine, and

reagents for LDH release assay were from Sigma (St. Louis, MO). Coomassie Plus

Protein Assay and SuperSignal West Pico chemiluminescent substrate were obtained

from Pierce (Rockford, IL). Fetal bovine serum, Dulbecco’s modified Eagle medium

(DMEM) containing 25 mM glucose, 1 mM sodium pyruvate, 19.4 µM pyridoxine hydrochloride, 2 mM glutamine, B-27 supplement were purchased from Invitrogen

(Carlsbad, CA). Tissue culture plates and pippets were manufactured by Sarstedt

(Newton, NC). The following antibodies were used in this study: mouse 15A3

antibody (dilution, 1:250; QED Bioscience, San Diego, CA), mouse anti-microtubule-

associated protein 2 mAb (MAP2, 1:1000; NeoMarkers, Fremont, CA), rabbit anti-

excitatory amino acid transporter 3 pAb (EAAT3, 1:1000; generously provided by Dr.

Jeffrey D. Rothstein), rabbit anti-glial fibrillary acidic protein pAb (GFAP, 1:1000;

Promega, Madison, WI), goat anti-β-actin mAb (1:4000; Santa Cruz Biotechnology;

Santa Cruz, CA) and rabbit anti-human Cu/Zn SOD1 pAb (1:2000; Santa Cruz

118

Biotechnology). Alexa Fluor 488 goat-anti-mouse or rabbit IgG, Alexa Fluor 594

goat-anti-mouse or rabbit IgG, and Hoechst 33342 were obtained from Molecular

Probes (Eugene, OR).

5.3.2 Primary rat cortical neuronal culture

Cerebral cortical neurons were prepared from embryonic day (E) 18-19 or

newborn rat pups (Meiners and Geller, 1997). The cortices were dissected out of the

brains, incubated in activated papain for 30 min at 37 °C, triturated by repeated

pipetting with a small bore pipette, and plated onto poly-D-lysine-coated (0.1 mg/ml)

plastic culture dishes or glass slides. These cultures were maintained in DMEM, 0.5%

fetal bovine serum and 1x B-27 supplement.

5.3.3 Immunofluorescence

The cultures were plated onto poly-D-lysine coated glass coverslips and fixed

with 2% paraformaldihyde supplemented with 1.7% sucrose in phosphate buffer for

30 min at room temperature. After permeabilized with 0.1% saponin in 1x PBS, the cultures were blocked with 1% bovine serus albumin (BSA) in 1x PBS with 0.1% saponin for 1 hr at room temperature and then incubated with primary antibodies overnight at room temperature. After rinsing, the coverslips were then incubated with secondary antibody solution (Alexa Fluor 488 goat-anti-mouse or rabbit IgG (1:1000),

and Alexa Fluor 594 goat anti-rabbit or mouse IgG (1:1000) in PBS with BSA) for 60

min at room temperatures. Nuclear staining was accomplished by adding Hoechst

119

33342 fluorescence dye into secondary antibody solution (0.25 µg/ml). The coverslips

were then rinsed and mounted onto glass slides with ImmuMount (Shandon Lipshaw).

5.3.4 Western blotting

Protein extracts were generated from primary cultures, resolved by SDS–PAGE and transferred onto PVDF membranes. After blocked by 5% non-fat milk in PBS

with 0.1% Tween-20, the membrane was incubated with the following primary

antibodies in 1% milk overnight at 4 ºC (rabbit anti-hSOD1 pAb or goat anti-β-actin

mAb) followed by peroxidase-conjugated goat anti-rabbit IgG (1:1200; ICN

Biomedicals, Aurora, OH) for 60 minutes at room temperature. The immunoreactive

bands were detected using the SuperSignal West Pico Chemiluminescent Substrate

according to the manufacturer's directions.

5.3.5 Isolation of total RNAs and messenger RNAs (mRNAs)

Total RNA was isolated from cultures using TRIZOL (Invitrogen) according to

manufacturer’s directions. Briefly, 1 ml of TRIZOL per well in 6-well plates was

added to cultures, and incubated at room temperature for 5 min. 0.2 volumes of

chloroform was added to the sample and mixed by vigorous shaking for 15 sec and

incubation at room temperature for 3 min. The organic phase was separated from the

aqueous phase by centrifugation at 12,000xg for 15 min at 4 oC. The aqueous phase

was transferred to a new tube and the RNA was precipitated with 2-propanol (0.5

volumes; 10 min at room temperature). The RNA was pelleted by centrifugation at

120

12,000xg for 10 min at 4 oC. The RNA pellet was then washed with 1 volume 70%

ethanol and pelleted by centrifugation at 7,500xg for 5 min at 4 oC. The RNA pellet

was air dried and then resuspended in 20-100 µl DEPC-treated ddH2O.

mRNA was isolated using the Oligo-tex mRNA Purification kit (Qiagen, Valencia,

CA). Briefly, 100µg of total RNA was mixed with 250µl buffer OEB and 15µl

Oligotex suspension and incubated at 70˚C for 5 min. The mixture was then incubated

in rotation at room temperature for another 20 min. The mRNA-Oligotex complex

was pelleted by centrifugation at 14,000 rpm for 3 min at room temperature. The

pellet was washed with buffer OW2 twice and pelleted by centrifugation at 14,000 rpm for 2 min at room temperature. mRNA was eluted off from Oligotex by incubating the pellet with buffer EB at 70˚C for 5 min. After centrifugation at 14,000 rpm for 2 min, the elution buffer containing mRNA was transferred to a new microtube and the concentration of mRNA was measured with the OD value at

260nm UV light by Beckman DU640 spectrophotometry (Beckman Instruments Inc,

Palo Alto, CA).

5.3.6 Immunoprecipitation of oxidized mRNAs

1.5 µg of mRNAs were incubated with 2.5 µg of anti-8OHG antibody 15A3 at room temperature for 1 hour. 20 µl of immobilized Protein L gel beads (PIERCE) were added and incubated at 4 oC for another 15 hours. The beads were washed three times by PBS with 0.04 % (v/v) NP-40 (Roche Applied Science, Indianapolis, IN).

After precipitation of beads by centrifuge, the oxidized mRNA:antibody:protein L

121

complexes were separated from non-oxidized mRNAs, which remained in the supernatant. Non-oxidized mRNAs were precipitated with ethanol and dissolved in

10 µl DEPC-treated H2O. The following items were added in order into the

precipitant of oxidized mRNAs: 300 µl of PBS with 0.04 % NP-40, 30 µl of 10 %

(w/v) sodium dodecyl sulfate (SDS), and 300 µl of PCI (phenol: chloroform: isoamyl alcohol; 25:24:1). The mixture was incubated at 37 oC for 15 minutes (vortexing every 5 minutes) and separated to aqueous phase and organic phase by spinning at

14,000 rpm for 5 minutes. The aqueous layer containing oxidized mRNAs was collected, and mixed with 40 µl of 3 M sodium acetate with pH 5.2, 2 µl of 5 mg/ml

glycogen, plus 1 ml of 95 % (v/v) ethanol. The sample was frozen at -80 oC for 1 hour

and centrifuged for 20 minutes. The pellet was washed by 75% ethanol and air-dried.

It was resuspended in 10 µl DEPC-treated H2O.

5.3.7 RT-PCR

mRNAs from cultures or immunoprecipitated mRNAs were reversely transcribed

with AMV reverse transcriptase (Invitrogen) and gene-specific primers, including

Cu/Zn SOD1 F (5’-atggcgacgaaggccgtgtgcgt) and R

(ctggcaaaatacaggtcattgaaacagaca), Glutamate dehydrogenase 1 F

(agttccaagacaggatatcgggt) and R (5’-actcagcttgtacatgatccgcta), β-actin F (5’-

cgggacctgacagactacctcat-3’) and R (5’-accgactgctgtcaccttcacc-3’). PCRs were

performed in the presence of 3 mM MgCl2, 0.2 mM dNTP, 0.25 µM primers and 2 U

Taq DNA polymerase (Invitrogen) in 1x PCR buffer. A series of cycles (20, 25, 30

122

cylces) were performed (95 °C for 30 s, specific annealing temperature to each set of

primers for 45 s, and 72 °C for 1 min). PCR products were visualized as single bands

on 1% agarose gels stained with ethidium bromide.

5.3.8 LDH release assay

Cytotoxicity was assessed by measuring the amount of lactate dehydrogenase (LDH) released from the cells following treatment. Aliquots of culture media were removed from wells at various time points after treatment. Culture media (50 µl/reaction) was mixed with an equal volume of substrate mixture (25 mM sodium lactate, 3.76 mM ß-

NAD+ (nicotinamide adenine dinucleotide), 603.4 µM MTT (3-[4,5-dimethylthiazol-

2-yl]-2,5-diphenyltetrazolium bromide), 100 µM MPMS (1-methoxyphenazine methosulfate) and 0.1% Triton X-100 in 200 mM Tris–HCl, pH 8.0) and incubated for

30 min at 37°C. The reaction was stopped by adding 100 µl stop solution (20% SDS

in 50% dimethylformamide). Absorbance readings were obtained with a GENios

microplate reader (Tecan Co., San Jose, CA, USA; ex=570 nm).

5.4 Results

5.4.1 The toxic effects of different insults associated with AD to neurons.

We applied rat neonatal primary neuronal cultures to examine the relationship of

RNA oxidation with neuronal cell death. Different insult conditions, including exposure of neurons to H2O2, glutamate, or amyloid β (Aβ) 25-35, were used to

generate cell models with oxidative stress. All these conditions have been implicated

123

in AD and mediate the production of ROS. Primary cultures contained less than 10% of astrocytes as demonstrated by immunofluorescence staining of GFAP. Cultures were treated with 50 µM H2O2 or 50 µM glutamate for 30 min followed by various recovery time periods, or 20 µM Aβ25-35 for 12 to 48 hr. The dosage of insults and the length of treatments were carefully selected based on literatures. The neuron-toxic effects of different insults were first examined with cell morphological changes observed by immunostaining of MAP2 (Figure 5.1). Neurons were most susceptible to the toxicity from the glutamate treatment. No intact cell morphology was observed under light microscopy 24 hr after exposure of cultures with glutamate. Therefore, immunoreactivity of MAP2 in glutamate-treated cultures was measured at the 10- hour time point. Under all three conditions, neuronal cells undergo degeneration as indicated by bead-like structures in neurites. The overall staining of MAP2 was decreased in oxidant-treated cultures compared to that in DMEM-treated controls.

The increased signal intensity observed in bead-like structures may result from the aggregation of MAP2 in degenerating or swelling neurites.

124

Figure 5.1 MAP2 immunostaining to examine neuron-toxic effects of different insults in primary neuronal cultures. Cultures were treated with 50 µM H2O2 or 50 µM glutamate for 30 min followed by selected recovery time, or 20 µM Aβ25-35 for 24 hr. The cultures were immunolabeled with monoclonal anti-MAP2 antibody. Alex 488 goat-anti- mouse IgG antibody was used for H2O2-treated cultures as shown in green fluorescence, and Alex 594 goat-anti-mouse IgG antibody was used for glutamate- or Aβ-treated cultures as shown in red fluorescence. 125

5.4.2 H2O2 treatment induces RNA oxidation in the early stage of neuronal cell death.

The purpose of this study was to investigate whether RNA oxidation occurs in

neuronal cultures under the insult of H2O2 and then the time course of RNA oxidation

in relation to that of cell death. Cultures were exposed to 50 µM H2O2 for 30 min

followed by selected recovery periods. RNA oxidation was detected by

immunolabeling with 15A3 antibody. Figure 5.2 showed that intensity of

fluorescence was increased 1 hr after treatment, continuously enhanced at 4 hr but

diminished at 10 hr time point, while no positive signal was detected in DMEM-

treated cultures, which were set as controls, after 4 hr incubation.

Figure 5.2 Immunostaining with 15A3 antibody to examine the time course of RNA oxidation in H2O2-treated neuronal cultures. Cultures were treated with 50 µM H2O2 for 30 min followed by selected recovery periods, including 1 hr, 4 hr and 10 hr. RNA oxidation was detected with 15A3 antibody 1 hr post-treatment, enhanced at 4 hr, but declined at the 10 hr time point. DMEM-treated cultures were used as controls. No significant increase of RNA oxidation was present in DMEM-treated cells at any selected time point. 4 hr treatment with DMEM was showed.

126

Nuclear staining was performed to examine cell apoptosis. The result showed

that no significant increase of apoptotic nuclei was observed within 15 hr after H2O2

treatment (Figure 5.3 A). Therefore, we selected a more sensitive method, LDH

release assay, to monitor the viability of treated cells (Figure 5.3 B). H2O2 treatment

caused time-dependent membrane damage in neuronal cultures as shown by LDH

release into the media. The H2O2-induced LDH release was not evident at 2 hr but

increased significantly by 6 hr, and then continuously elevated with longer incubation

time. The conclusion that RNA oxidation is an early event in the process of cell death

was made based on the following facts: (1) the immunofluorescence from oxidized

RNA was increased by 4 hr after treatement but declined at the 10 hr time point; (2) a

gradual increase of LDH release value was found within 15 hr post-treatment; (3) no

significant increase of apoptotic nuclei was detected 15 hr after treatment.

To examine whether mRNA is the group of RNA oxidized and further identify

the oxidized mRNA species in cultures, we performed immunoprecipitation to isolate

oxidized mRNAs. DIG labeling on reversely transcribled cDNA from oxidized

mRNA was used to facilitate the visualization of oxidized mRNAs by Southern blotting. Figure 5.4 showed that a significant time-dependent increase of the level of oxidized mRNAs was observed within 4 hr after treatment as compared to non-treated control.

127

Figure 5.3 Examination of cell viability by nuclear staining and LDH release assay. Cultures were treated with 50 µM H2O2 for 30 min followed by selected recovery periods. No increase of apoptotic nuclei was detected in H2O2-treated culture 15 hr post-treatment (A). LDH release assay showed a time-dependent increase of LDH release into media after treatment.

128

Figure 5.4 Immunoprecipitation of oxidized mRNAs in H2O2-treated neuronal cultures. Cultures were exposed to 50 µM H2O2 for 30 min followed by selected recovery periods. A time-dependent increase of the level of oxidized mRNAs was observed within 4 hr post-treatment.

5.4.3 RNA oxidation is increased in neuronal cultures treated with glutamate and

Aβ25-35.

In addition to H2O2 treatment, we also examined RNA oxidation in primary

neuronal cultures after exposure to glutamate and Aβ25-35. Cultures were incubated

with 50 µM glutamate for 30 min followed by selected recovery periods or 20 µM

Aβ25-35 for certain periods. RNA oxidation was indicated by immunofluorescence

with 15A3 antibody, while cell viability was monitored by LDH release assay. As

shown in Figure 5.5, a significant increase of immunolabeling of RNA oxidation was

detected as early as 4 hr after glutamate treatment (A) and 12 hr after Aβ25-35

treatment (B). LDH release assay showed an increase of LDH release into media,

which became evident after RNA oxidation was observed.

129

Figure 5.5 RNA oxidation in glutamate- or Aβ-treated neuronal cultures. Cultures were exposed to 50 µM glutamate for 30 min followed with selected recovery periods or 20 µM Aβ for 12, 24 and 36 hrs. RNA oxidation was detected in treated cultures in the early stage of cell damage which was monitored by LDH release assay.

130

5.4.4 RNA oxidation is primarily observed in neurons, not in astrocytes

An important question motivating this study is what group of cells is more susceptible to RNA oxidation. Here we investigated this issue by localizing oxidized

RNA in different cell types, neurons and astrocytes. We used the antibody against

excitatory amino acid transporter 3 (EAAT3), which is mainly located on neuronal

cells, to specify neurons. GFAP immunolabeling was used to identify astrocytes.

Results on double-labeling of 15A3 with anti-EAAT3 or anti-GFAP antibody showed

that oxidized RNA was primarily located in certain groups of neuronal cells (Figure

5.6).

Figure 5.6 Localization of oxidized RNAs in neuronal cells. Cultures were treated with 50 µM H2O2 for 30 min followed with additional 4 hr recovery period. Double-immunolabeling of 15A3 with EAAT3, a neuron-specific protein, or GFAP, an astrocyte marker, was performed to localized oxidized RNAs in specific cell types. The result showed that oxidized RNAs are primarily present in neurons but not astrocytes.

131

5.4.5 Identification of oxidized mRNA species and further investigation of possible consequence of oxidized mRNA species in H2O2-treated neuronal cultures.

Oxidized mRNA species in AD brain were identified in our previous study

(Chapter 3). Whether neuron cultures under oxidative stress have the same profile of oxidized mRNA species as that in AD brain is important to know because it will validate the applicability of neuronal cultures to study the consequence of oxidized mRNAs. We used immunoprecipitation method to isolate oxidized mRNAs from the neuronal culture treated with 50 µM H2O2 for 30 min followed by a 4 hr recovery period. mRNAs were harvested from cultured cells and oxidized mRNAs were immunoprecipitated using 15A3 antibody. RT-PCR of some mRNA species, which were highly oxidized in AD brains, was performed in oxidized mRNA pool of the culture. As shown in Figure 5.7, human Cu/Zn SOD1 (hSOD1) and glutamate dehydrogenase 1 (GDH) were found oxidized in the cultures. β-actin, which is a highly abundant mRNA species and is not oxidized in AD brains, was also not detected in oxidized mRNA pool from cultures. This indicated that primary neuronal culture may use the same mechanism to selectively oxidize mRNAs as the neurons in the AD brain.

132

Figure 5.7 Identification of oxidized mRNA species in H2O2-treated neuronal cultures by RT-PCR. Cultures were treated with 50 µM H2O2 (+) or DMEM (–) for 30 min followed with an additional 4 hr recovery period. mRNAs were harvested for immunoprecipitation of oxidized mRNAs with 15A3 antibody. RT- PCR was performed to examine human Cu/Zn SOD1 (hSOD1), glutamate dehydrogenase 1 (GDH) and β-actin mRNA species in oxidized mRNA pools. hSOD1 and GDH but not β-actin mRNAs were oxidized in H2O2-treated neuronal cultures. RT-PCR on mRNAs isolated from cultures was used as controls (ctrl).

We observed the decrease of normal protein expression from oxidized mRNAs in our previous in-vitro study (chapter 4). Whether it occurs to oxidized mRNAs in primary cultures is the next issue to study. Western blotting analysis was performed to examine the protein levels of Cu/Zn SOD1, which represented oxidized mRNAs in the cultures, and β-actin, which is not oxidized (Figure 5.8 A). Cu/Zn SOD1 content remained unchanged by 6 hr but was dramatically decreased at 12 hr post-treatment, while β-actin had the same expression levels at these different time points. To examine whether mRNA levels of Cu/Zn SOD1 and β-actin were altered under oxidative stress, we did RT-PCR using mRNAs isolated from the cultures. Figure 5.8

B showed that the levels of both Cu/Zn SOD1 and β-actin mRNAs were not different in non-treated and treated cultures. This suggested that the decrease of protein expression was regulated at post-transcriptional level. Therefore, mRNA oxidation can be considered as a possible contributor to the decreased translational efficiency.

133

Figure 5.8 Examination of the protein and mRNA levels of oxidized mRNA species in neuronal cultures. Cultures were exposed to 50 µM H2O2 (+) or DMEM (–) for 30 min followed with 3 hr, 6 hr and 12 hr recovery periods. The protein levels of hSOD1 and β-actin were examined by Western blotting analysis (A). hSOD1 protein expression was decreased 12 hr after H2O2 treatment, while β-actin remained unchanged. mRNA levels were examined by RT-PCR using mRNAs isolated from cultures after treatment (B). No difference of mRNA levels was detected in H2O2- or DMEM-treated cultures even at 12 hr time point. Different PCR cycles were performed for quantitative comparison.

5.5 Discussion

Numerous studies have been performing to investigate the potential pathogenic effects of oxidized macromolecules in neurodegeneration. In our previous study

(chapter 2) we found that 30-70% of mRNAs are oxidized in Alzheimer brains. To better understand the involvement of oxidized RNA in oxidative stress-related cell

134

death, we investigated the time courses of RNA oxidation and cell damage as well as

the localization of oxidized RNA using primary neuronal cultures. H2O2, glutamate

and Aβ, all of which have been proposed to mediate oxidative stress and associate

with cell death in AD, were applied to neuronal cultures to generate cell models with

oxidative insults. H2O2 is normally generated in cells via dismutation of the

superoxide anion that is a byproduct of mitochondrial respiratory chain. H2O2 itself

has relatively low oxidative activity and is disposed by antioxidant enzymes including

catalase and glutathione peroxidase. Mitochondrial dysfunction and inflammatory

response from microglial cells in AD can enhance the production of H2O2, which is

freely diffusible and can be converted to the highly reactive hydroxyl radical in the presence of iron. In our present study, 50 µM H2O2 did not cause apparent cell

damage until 4 hr after exposure based on the LDH release assay. The increase of

oxidized RNA level was detectable as early as 1 hr post-treatment with

immunofluorescence and immunoprecipitation (Figures 5.2 and 5.4). There is a

time-dependent rise in the level of RNA oxidation within 4 hr. The initial oxidative

modification on RNA may be the result of the attacks from intracellular ROS

generated directly from diffused H2O2. Cells possess a certain level of oxidative

capacity and dispose oxidants via several pathways. Once the capacity is

overwhelmed, a chain reaction producing ROS may be turned on intracellularly in a

positive-feedback way. This may explain the rise of oxidized RNA levels within 4 hr

after exposing cells with H2O2. However, the level of RNA oxidation did not keep increasing as to the level of cell damage. By 15 hr we visualized the decline of

135

oxidative damage on RNA as compared to that at 4 hr (Figure 5.2). This indicates that RNA oxidation is an early event in the process of cell death induced by oxidative insults.

Glutamate is involved in the majority of excitatory neurotransmission. Excessive exposure of glutamate to cells can generate oxidative toxicity that has been reported by several research groups (Schubert and Piasecki, 2001). 10 min treatment with 30

µM NMDA to neuronal culture can induce a striking increase of intracellular ROS concentration (Bindokas et al., 1996). The ability of Aβ on regulating the production of ROS has been also discovered by numerous studies. Aβ facilitates ROS generation by multiple mechanisams: (1) Aβ can cause dysfunction of mitochondria which is the major site to produce free radicals (Casley et al., 2002); (2) Aβ can impair intracellular antioxidant systems (Kim et al. 2003); (3) Aβ can cause the accumulation of extracellular glutamate by inhibition of glutamate transporters

(Facheris et al. 2004). Therefore, neuronal cultures treated with glutamate or Aβ can produce the phenomena of oxidative stress and are good cell models to investigate

RNA oxidation. Consistent with the result from H2O2-treated cultures, we found RNA oxidation occurs in the early stage of cell death in glutamate- or Aβ-treated cultures.

Nunomura et al. (1999) observed that increased RNA oxidation in AD was restricted to healthy-looking vulnerable neurons. It was not dependent on proximity to senile plaques and was reduced in neurons containing neurofibrillary tangles. This

136

suggested that RNA oxidation is an early event preceding cell death in AD. Our lab has also examined RNA oxidation in transgenic mice expressing mutant SOD1

(G93A) at different stages of disease. The SOD1 (G93A) mouse develops a neurological disorder that resembles amyotrophic lateral sclerosis (ALS) (Gurney et al., 1994). We observed significantly increased RNA oxidation in motor neurons of spinal cord at presymptomatic stage (2-month-old) when the motor neurons are still healthy (Chang et al., 2004). These motor neurons then gradually degenerated during the symptomatic stage (3- to 4-month-old). Additionally, we have investigated whether RNA oxidation occurs after spinal cord injury in a rat model. Post-traumatic necrosis occurs after spinal cord injury in the rat; apoptotic cells are additionally found from 6 hr to 3 weeks after injury especially in the spinal white matter (Crowe et al., 1997). These apoptotic cells are primarily oligodendrocytes. We observed significantly increased RNA oxidation localized in oligodendrocytes in the dorsal column rostral to lesion site at 1 day post-injury and these oligodendrocytes died 8-14 days later (Sun et al., 2003). All of these results suggest that RNA oxidation is not just a part of the consequence of cell death but is an early event preceding cell death.

This implicates the possibility of the involvement of RNA oxidation in the pathogenesis of AD.

In situ detection of oxidized RNAs showed that oxidized RNA is a prominent feature in vulnerable neurons in AD brains (Nunomura et al., 1999). However, the question remains unsolved that whether neuron is the specific group of cells

137

susceptible to RNA oxidation or accumulation of oxidized RNAs over decades occurs to neurons because they are non-dividing. We investigated this issue in primary cultures in which neurons and astrocytes were under the same level of oxidative insults. Double-labeling with 15A3 and cell type-specific markers was employed to localize oxidized RNA. Oxidized RNA is predominantly present in neurons identified by EAAT3 immunoreactivity, not in astrocytes identified by GFAP staining. A number of studies showed that astrocytes are the primary site of production of ROS that causes delayed neuronal death but has no effect on astrocytes themselves

(Abranov AY, 2004). The resistance of astrocytes to ROS has been ascribed to their greater antioxidant capacity as compared to neurons. The concentration of reduced glutathione, the major intracellular antioxidant, had been found to be twice as high in astrocytes as in neurons. Glutathione reacts directly with the radicals in nonenzymic reactions and is also an electron donor in the GPx-catalyzed reduction of peroxides.

hSOD1 mRNA, highly oxidized in AD brain, is also present in the oxidized mRNA pool in oxidant-insulted neuronal cultures. Based on the observation that

RNA oxidation impairs protein translational efficiency, we hypothesized and showed that hSOD1 protein level was decreased in neuronal cultures under oxidative damage.

By examining the hSOD1 mRNA level with RT-PCR, we eliminated the possibility that the decreased protein expression was derived from a decrease of hSOD1 mRNA transcripts in quantity. The protein level of β-actin, of which the mRNA species is not detected in oxidized mRNA pool in cultures, was not changed in oxidant-treated cells.

138

So we propose that the decrease of protein expression in some mRNA species may be

associated with RNA oxidation in AD brains, which is a possible pathway of how

RNA oxidation is involved in the pathogenesis of AD

139

CHAPTER 6

CONCLUSION AND PERSPECTIVES

Alzheimer’s disease (AD) is the most common cause of adult dementia. It is

characterized by the accumulation of senile plaques and neurofibrillary tangles and

neuronal cell loss in the brain. The pathogenesis of AD remains largely unknown although it has been extensively studied. Oxidative stress is believed to be an important contributor to the disease. Oxidative modifications to vital cellular components have been described in association with the disease. Cytoplasmic RNA oxidation is a prominent feature of vulnerable neurons in AD affected regions. This initiates several interesting questions: (1) which group of RNAs is oxidized; (2) what is the magnitude of oxidized RNAs; (3) whether oxidation to mRNAs is a selective or random event if mRNAs are oxidized, (4) which mRNA species are oxidized, (5) what is the consequence of mRNA oxidation, (6) whether mRNA oxidation is a simple result or a possible contributor of cell death. This dissertation investigates these issues in AD tissues and neuronal cultures under oxidative stress.

140

RNA oxidation occurs to a great extent in neurons that are especially vulnerable to degeneration in AD affected regions. In this dissertation, the magnitude of oxidized mRNAs in AD frontal cortexes was characterized as well as the oxidized mRNA species. 30-70% of mRNAs are oxidized in the frontal cortexes of AD as compared to age-matched normal controls. Identification of oxidized mRNA species showed that mRNA oxidation is not a random but selective event in AD. Many identified oxidized mRNA species have been implicated in the pathogenesis of AD. Quantitative analysis revealed that some mRNA species are more susceptible to oxidative damage and the relative amounts of oxidized transcripts reach 50 to 70% for some mRNA species.

Such high degree of oxidation on mRNAs suggests the importance of mRNA oxidation as a crucial event involved in AD. Therefore the study on the consequence of oxidized mRNAs became critical to understand the potential pathways by which oxidized mRNAs may have an impact in AD pathogenesis.

mRNA oxidation may have deleterious effects by interference with normal translational process or by a toxic gain-of-function itself. Examination of the protein expression from oxidized mRNAs revealed that mRNA oxidation leads to the decrease of normal protein level and its activity. The impairment of protein translation may result from polyribosome stalling on oxidized mRNA transcripts.

Microinjection of 0.125 pg or more of oxidized mRNAs into HT4 cells lead to cell death which can be mediated by oxidized mRNAs themselves or their translational consequence.

141

RNA oxidation occurs in the early stage of cell death in primary neuronal cultures, indicating that it is not a simple byproduct but a potential contributor in the cell death pathway. The preference of RNA oxidation to neurons rather than astrocytes also implicates its potential role in neurodegeneration. Some mRNA species highly oxidized in AD brains are also present in oxidized mRNA pools of neuronal cultures with oxidative stress. The investigation of the translational consequence of oxidized mRNA species in cultures showed the decrease of protein expression level.

There is a long incubation period between the initiation of intracellular aberrance and the development of symptoms of AD. How oxidative stress develops at the presymptomatic stage of the disease remains to be resolved. It is believed that the consequent oxidation to intracellular molecules mediates the toxic effects to neurons in the disease. Because the progressive course of AD is so long during which many impact factors may interact to alter the ultrastructure and architecture inside the neurons, the AD brain tissue is not a suitable model to study the consequence of mRNA oxidation. However, study of the consequence of mRNA oxidation in vivo is required to directly link it with the pathogenesis of the disease. The levels and functional activities of proteins translated from oxidized mRNA species can be examined. The possibility that oxidized mRNAs generate mutant and/or truncated proteins can not be excluded from the studies in this dissertation. Protein sequencing analysis by Tandom Mass Spectrometry (MS-MS) can be applied to investigate this

142

issue. The proteins translated from oxidized mRNA species can also be examined

with mutations or truncations using AD postmortem tissues. Blockage of reactive

oxygen species generation by supplying antioxidants to AD subjects or animal

models or other mechanisms may help to reveal the consequence of mRNA oxidation

if the treatment can rescue the neurons from oxidative stress. This result can not be

conclusive because inhibition of oxidation to other molecules may be involved in the

protective effects. The better strategy to study the consequence of oxidized mRNAs is to block or repair oxidation to mRNAs specifically or to degrade oxidized mRNAs. In this dissertation, it is showed that mRNA oxidation is a selective event in AD. The knowledge about the mechanism of this selectivity, which is currently unknown, will help to design a way to specifically block mRNA oxidation. RNA repair system has long been underestimated. However, the recent discovery of RNA repair enzyme rectified the traditional view and RNA repair system is still an uncharacterized field.

RNA repair enzyme for oxidation may be present in cells since ROS generation and nucleic acid oxidation occur in normal cellular metabolism. Its discovery, which is a long way to go, will be essential to understand the consequence of RNA oxidation.

This dissertation suggests that RNA oxidation may play a more important role in

AD pathogenesis than previously anticipated, and that any therapeutic intervention targeting at oxidative damages to intracellular molecules initiated at the earliest possible stage of the disease may be benefiicial.

143

LIST OF REFERENCES

Aas PA, Otterlei M, Falnes PO, Vagbo CB, Skorpen F, Akbari M, Sundheim O, Bjoras M, Slupphaug G, Seeberg E, Krokan HE (2003) Human and bacterial oxidative demethylases repair alkylation damage in both RNA and DNA. Nature 421:859-63.

Abe T, Tohgi H, Isobe C, Murata T, Sato C (2002) Remarkable increase in the concentration of 8-hydroxyguanosine in cerebrospinal fluid from patients with Alzheimer's disease. J Neurosci Res. 70:447-50.

Abramov AY, Canevari L, Duchen MR (2004) β-amyloid peptides induce mitochondrial dysfunction and oxidative stress in astrocytes and death of neurons through activation of NADPH oxidase. J Neurosci 24:565-575.

Adamec E, Mohan P, Vonsattel JP, Nixon RA (2002) Calpain activation in neurodegenerative diseases: confocal immunofluorescence study with antibodies specifically recognizing the active form of calpain 2. Acta Neuropathol 104:92-104.

Aksenov MY, Aksenova MV, Butterfield DA, Geddes JW, Markesbery WR (2001) Protein oxidation in the brain in Alzheimer's disease. Neuroscience.103:373-83.

Almeida A, Delgado-Esteban M, Bolanos JP, Medina JM (2002) Oxygen and glucose deprivation induces mitochondrial dysfunction and oxidative stress in neurons but not in astrocytes in primary culture. J Neurochem. 81:207-217.

Amici A, Levine RL, Stadtman ER (1989) Conversion of amino acids residues in proteins and amino acid homopolymers to carbonyl derivatives by metal-catalyzed reactions. J. Biol. Chem. 264:3341-3346.

Andreasen N, Minthon L, Davidsson P, Vanmechelen E, Vanderstichele H, Winblad B, Blennow K (2001) Evaluation of CSF-tau and CSF-Aβ42 as diagnostic markers for Alzheimer’s disease, Arch. Neurol. 58:373–379.

Bindokas VP, Jordan J, Lee CC, Miller RJ (1996) Superoxide production in rat hippocampal neurons: selective imaging with hydroethidine. J Neurosci. 16:1324- 1336.

Andersen JK (2004) Oxidative stress in neurodegeneration: cause or consequence? Nat. Rev. Neuroscience S18-S25.

144

Arendt T, Holzer M, Grossmann A, Zedlick D, Bruckner MK (1995) Increased expression and subcellular translocation of the mitogen activated protein kinase kinase and mitogen-activated protein kinase in Alzheimer's disease. Neuroscience 68:5-18.

Arispe N, Pollard HB, Rojas E (1994) beta-Amyloid Ca(2+)-channel hypothesis for neuronal death in Alzheimer disease. Mol. Cell Biochem. 140:119-25.

Arnold SE, Hyman BT, Flory J, Damasio AR, Van Hoesen GW (1991) The topographical and neuroanatomical distribution of neurofibrillary tangles and neuritic plaques in the cerebral cortex of patients with Alzheimer's disease. Cereb Cortex. 1:103-16.

Balcz B, Kirchner L, Cairns N, Fountoulakis M, Lubec G (2001) Increased brain protein levels of carbonyl reductase and alcohol dehydrogenase in Down syndrome and Alzheimer's disease. J Neural Transm Suppl 193-201.

Barnham KJ, Masters CL, Bush AI (2004) Neurodgenerative diseases and oxidative stress. Nat Rev. Drug Discovery 3:205-214.

Belloir B, Kovari E, Surini-Demiri M, Savioz A (2001) Altered apolipoprotein D expression in the brain of patients with Alzheimer disease. J Neurosci Res 64:61-69.

Behl C, Davis JB, Lesley R, Schubert D (1994) Hydrogen peroxide mediates amyloid β protein toxicity. Cell 77:817-827.

Berardi P, Russell M, El-Osta A, Riabowol K (2004) Functional links between transcription, DNA repair and apoptosis. Cell Mol Life Sci. 61:2173-80.

Berens C, Streicher B, Schroeder R, Hillen W (1998) Visualizing metal-ion-binding sites in group I introns by iron(II)-mediated Fenton reactions. Chem Biol. 5:163-75.

Berlett BS, Levine RL, Stadtman ER (1998) Carbon dioxide stimulates peroxynitrite- mediated nitration of tyrosine residues and inhibits oxidation of methionine residues of glutamine synthetase: both modifications mimic effects of adenylylation. Proc Natl Acad Sci U S A. 95:2784-9.

Blass JP, Baker AC, Ko LW, Black RS (1990) Induction of Alzheimer Antigens by an uncoupler of oxidative phosphorylation. Arch Neurol 47:864–869.

Blass JP and Gilson GE (1991) The role of oxidative abnormalities in the pathophysiology of Alzheimer’s disease. Rev Neurol (Paris) 147:513-525.

145

Boje KM and Arora PK (1992) Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death. Brain Res 587:250-256.

Bowes MP, Masliah E, Otero DA, Zivin JA, Saitoh T (1994) Reduction of neurological damage by a peptide segment of the amyloid beta/A4 protein precursor in a rabbit spinal cord ischemia model. Exp Neurol. 129:112-9.

Bowling AC and Beal MF (1995) Bioenergetic and oxidative stress in neurodegenerative diseases. Life Science 56:1151-1171.

Browne SE, Bowling AC, MacGarvey U, Baik MJ, Berger SC, Muqit MMK, Bird ED, Beal MF (1997) Oxidative damage and metabolic dysfunction in Huntington’s disease: Selective vulnerability of the basal ganglia. Ann. Neurol. 41:646–653.

Browne SE, Ferrante RJ, Beal MF (1999) Oxidative stress in Huntington's disease. Brain Pathol. 9:147-63.

Butterfield DA, Castegna A, Lauderback CM, Drake J (2002) Evidence that amyloid beta-peptide-induced lipid peroxidation and its sequelae in Alzheimer's disease brain contribute to neuronal death. Neurobiol Aging. 23:655-64.

Butterfield DA and Lauderback CM (2002) Lipid peroxidation and protein oxidation in Alzheimer’s disease brain: potential causes and consequences involving amyloid beta-peptide associated free radical oxidative stress. Free Rad. Biol. Med. 32:1050- 1060.

Carri MT, Ferri A, Cozzolino M, Calabrese L, Rotilio G (2003) Neurodegeneration in amyotrophic lateral sclerosis: the role of oxidative stress and altered homeostasis of metals. Brain Res Bull. 61:365-74

Casley CS, Land JM, Sharpe MA, Clark JB, Duchen MR, Canevari L (2002) Beta- amyloid fragment 25-35 causes mitochondrial dysfunction in primary cortical neurons. Neurobiol Dis. 10:258-67.

Castegna A, Aksenov M, Aksenova M, Thongboonkerd V, Klein JB, Pierce WM, Booze R, Markesbery WR, Butterfield DA (2002a) Proteomic identification of oxidatively modified proteins in Alzheimer's disease brain. Part I: creatine kinase BB, glutamine synthase, and ubiquitin carboxy-terminal hydrolase L-1. Free Radic Biol Med. 33:562-71.

Castegna A, Aksenov M, Thongboonkerd V, Klein JB, Pierce WM, Booze R, Markesbery WR, Butterfield DA (2002b) Proteomic identification of oxidatively modified proteins in Alzheimer's disease brain. Part II: dihydropyrimidinase-related protein 2, alpha-enolase and heat shock cognate 71. J Neurochem. 82:1524-32.

146

Castegna A, Thongboonkerd V, Klein JB, Lynn B, Markesbery WR, Butterfield DA (2003) Proteomic identification of nitrated proteins in Alzheimer's disease brain. J. Neurochem. 85:1394-401.

Chan SL, Mattson MP (1999) Caspase and calpain substrates: roles in synaptic plasticity and cell death. J Neurosci Res 58:167-190.

Chang Y, Shan X, Lin CL (2004) RNA oxidation is an early event preceding motor neuron death in ALS. Soc. Neurosci. Abs. 96.3.

Cheng KC, Cahill DS, Kasai H, Nishimura S, Loeb LA (1992) 8-Hydroxyguanine, an abundant form of oxidative DNA damage, causes G→T and A→C substitutions. J. Biol. Chem. 267:166-172.

Colton CA, Gilbert DL (1987) Production of superoxide anions by a CNS macrophage, the microglia. FEBS Lett 223:284-288

Combs CK, Johnson DE, Cannady SB, Lehman TM, Landreth GE (1999) Identification of microglial signal transduction pathways mediating a neurotoxic response to amyloidogenic fragments of beta-amyloid and prion proteins. J Neurosci 19:928-939.

Combs CK, Karlo JC, Kao SC, Landreth GE (2001) -Amyloid stimulation of microglia and monocytes results in TNFalpha-dependent expression of inducible nitric oxide synthase and neuronal apoptosis. J Neurosci 21:1179-1188.

Connor JR, Menzies SL, St. Martin SM, Mufson EJ (1992) A histochemical study of iron, transferrin, and ferritin in Alzheimer’s disease brains. J. Neurosci. Res. 31:75-83.

Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell PC, Small GW, Roses AD, Haines JL, Pericak-Vance MA (1993) Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science. 261:921-3.

Crowe MJ, Bresnahan JC, Shuman SL, Masters JN, Beattie MS (1997) Apoptosis and delayed degeneration after spinal cord injury in rats and monkeys. Nat Med. 3:73-6.

Davies E, Abe S (1995) Methods for isolation and analysis of polyribosomes. Methods Cell Biol. 50:209-22.

Davydov V, Hansen LA, Shackelford DA (2003) Is DNA repair compromised in Alzheimer's disease? Neurobiol Aging. 24:953-68.

147

Dean RT, Fu s, Stocker R, Davies MJ (1997) Biochemistry and pathology of radical- mediated protein oxidation. Biochem. J. 324:1-18.

Dexter DT, Carter CJ, Wells FR, Javoy-Agid F, Agid Y, Lees A, Jenner P, Marsden CD (1989) Basal lipid peroxidation in substantia nigra is increased in Parkinson's disease. J. Neurochem. 52, 381–389

Dickey LF, Petracek ME, Nguyen TT, Hansen ER, Thompson WF (1998) Light regulation of Fed-1 mRNA requires an element in the 5' untranslated region and correlates with differential polyribosome association. Plant Cell. 10:475-84.

Drake J, Petroze R, Castegna A, Ding Q, Keller JN, Markesbery WR, Lovell MA, Butterfield DA (2004) 4-Hydroxynonenal oxidatively modifies histones: implications for Alzheimer's disease. Neurosci Lett. 356:155-8.

Eberwine J, Yeh H, Miyashiro K, Cao Y, Nair S, Finnell R, Zettel M, Coleman P (1992) Analysis of gene expression in single live neurons. Proc Natl Acad Sci U S A 89:3010-3014.

Estevez AG, Crow JP, Sampson JB, Reiter C, Zhuang Y, Richardson GJ, Tarpey MM, Barbeito L, Beckman JS (1999) Induction of nitric oxide-dependent apoptosis in motor neurons by zinc-deficient superoxide dismutase. Science 286:2498-500.

Facheris M, Beretta S, Ferrarese C (2004) Peripheral markers of oxidative stress and excitotoxicity in neurodegenerative disorders: tools for diagnosis and therapy? J Alzheimers Dis. 6:177-84.

Ferrer I, Blanco R, Carmona M, Ribera R, Goutan E, Puig B, Rey MJ, Cardozo A, Vinals F, Ribalta T (2001) Phosphorylated map kinase (ERK1, ERK2) expression is associated with early tau deposition in neurones and glial cells, but not with increased nuclear DNA vulnerability and cell death, in Alzheimer disease, Pick's disease, progressive supranuclear palsy and corticobasal degeneration. Brain Pathol 11:144- 158.

Fiala ES, Conaway CC, Mathis JE (1989) Oxidative DNA and RNA damage in the livers of Sprague-Dawley rats treated with the hepatocarcinogen 2-nitropropane. Cancer Res. 49:5518-22.

Fisher N, Meunier B (2001) Effects of mutations in mitochondrial cytochrome b in yeast and man. Deficiency, compensation and disease. Eur J Biochem 268:1155-1162.

Focht SJ, Snyder BS, Beard JL, Van Gelder W, Williams LR, Connor JR (1997) Regional distribution of iron, transferrin, ferritin, and oxidatively-modified proteins in young and aged Fischer 344 rat brains. Neuroscience 79:255-261.

148

Furlong CE (2000) PON1 status and neurologic symptom complexes in Gulf War veterans. Genome Res 10:153-155.

Gabbita SP, Lovell MA, Markesbery WR (1998) Increased nuclear DNA oxidation in the brain in Alzheimer's disease. J Neurochem. 71:2034-40.

Gartner U, Holzer M, Arendt T (1999) Elevated expression of p21ras is an early event in Alzheimer's disease and precedes neurofibrillary degeneration. Neuroscience 91:1- 5.

Gasser T (2001) Genetics of Parkinson’s disease. J. Neurol. 248:833–840.

Ghadge GD, Lee JP, Bindokas VP, Jordan J, Ma L, Miller RJ, Roos RP (1997) Mutant superoxide dismutase-1-linked familial amyotrophic lateral sclerosis: molecular mechanisms of neuronal death and protection. J Neurosci. 17:8756-66.

Ghosh SS, Swerdlow RH, Miller SW, Sheeman B, Parker WD Jr, Davis RE (1999) Use of cytoplasmic hybrid cell lines for elucidating the role of mitochondrial dysfunction in Alzheimer’s disease and Parkinson’s disease. Ann N Y Acad Sci 893:176-191.

Giasson BI, Duda JE, Murray IV, Chen Q, Souza JM, Hurtig HI, Ischiropoulos H, Trojanowski JQ, Lee VM (2000) Oxidative damage linked to neurodegeneration by selective -synuclein nitration in synucleinopathy lesions. Science 290:985–989.

Goda Y (2002) Cadherins communicate structural plasticity of presynaptic and postsynaptic terminals. Neuron 35:1-3.

Good PF, Hsu A, Werner P, Perl DP, Olanow CW (1998) Protein nitration in Parkinson's disease. J. Neuropathol. Exp. Neurol. 57:338–342.

Good PF, Perl DP, Bierer LM, Schmeidler J (1992) Selective accumulation of aluminum and iron in the neurofibrilary tangles of Alzheimer’s disease: A laser microscope (LAMMA) study. Ann Neurol. 31:286-292

Goodman L (1953) Alzheimer’s disease: A clinico-pathologic analysis of twenty- three cases with a theory on pathogenesis. J. Nerv. Ment. Dis. 117:97-130.

Gotz ME, Double K, Gerlach M, Youdim MB, Riederer P (2004) The relevance of iron in the pathogenesis of Parkinson's disease. Ann N Y Acad Sci. 1012:193-208.

Gracy RW, Talent JM, Kong Y, Conrad CC (1999) Reactive oxygen species: the unavoidable environmental insult? Mutation research 428:17-22

149

Green DR, Reed JC (1998) Mitochondria and apoptosis. Science 281:1309–1312. Grundke-Iqbal I, Fleming J, Tung YC, Lassmann H, Iqual K, Joshi JG (1990) Ferritin is a component of the neuritic (senile) plaque in Alzheimer dementia. Acta Neuropathol. 81:105-110.

Gurney ME, Pu H, Chiu AY, Canto MCD, PolchowCY, Alexander DD, Caliendo J, Hentati A, Kwon YW, Deng HX, Chen W, Zhai P, Sufit RL, Siddique T (1994) Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 264:1772-5.

Harris ME, Hensley K, Butterfield DA, Leedle RA, Carney JM (1995) Direct evidence of oxidative injury produced by the Alzheimer’s β-amyloid peptide (1-40) in cultured hippocampal neurons. Exp. Neurol. 131:193-202.

Hasegawa E, Takeshige K, Oishi T, Murai Y, Minakami S (1990) 1-Methyl-4- phenylpyridinium (MPP+) induces NADH-dependent superoxide formation and enhances NADH-dependent lipid peroxidation in bovine heart submitochondrial particles. Biochem. Biophys. Res. Commun. 170:1049-1055.

Hartmann T, Bieger SC, Bruhl B, Tienari PJ, Ida N, Allsop D, Roberts GW, Masters CL, Dotti CG, Unsicker K, Beyreuther K (1997) Distinct sites of intracellular production for Alzheimer's disease A beta40/42 amyloid peptides. Nat Med. 3:1016- 20.

Hashimoto M, Rockenstein E, Crews L, Masliah E (2003) Role of protein aggregation in mitochondrial dysfunction and neurodegeneration in Alzheimer's and Parkinson's diseases. Neuromolecular Med. 4:21-36

Hayakawa H, Uchiumi T, Fukuda T, Ashizuka M, Kohno K, Kuwano M, Sekiguchi M (2002) Binding capacity of human YB-1 protein for RNA containing 8-oxoguanine. Biochemistry 41:12739–12744.

Henle ES, Han Z, Tang N, Rai P, Luo Y, Linn S (1999) Sequence-specific DNA cleavage by Fe2+-mediated fenton reactions has possible biological implications. J Biol Chem. 274:962-71.

Hensley K, Carney JM, Mattson MP, Aksenova M, Harris M, Wu JF, Floyd RA, Butterfield DA (1994) A model for β-amyloid aggregation and neurotoxicity based on free radical generation by the peptide: relevance to Alzheimer’s disease. Proc. Natl. Acad. Sci. USA 91:3270-3274.

150

Hensley K, Hall N, Subramaniam R, Cole P, Harris M, Aksenov M, Aksenova M, Gabbita SP, Wu JF, Carney JM, et al. (1995) Brain regional correspondence between Alzheimer's disease histopathology and biomarkers of protein oxidation. J Neurochem. 65:2146-56.

Hensley K, Maidt ML, Yu Z, Sang H, Markesbery WR, Floyd RA (1998) Electrochemical analysis of protein nitrotyrosine and dityrosine in the Alzheimer brain indicates region-specific accumulation. J Neurosci. 18:8126-32.

Higgins CM, Jung C, Xu Z (2003) ALS-associated mutant SOD1G93A causes mitochondrial vacuolation by expansion of the intermembrane space and by involvement of SOD1 aggregation and peroxisomes. BMC Neurosci. 4:16.

Hirai K, Aliev G, Nunomura A, Fujioka H, Russell RL, Atwood CS, Johnson AB, Kress Y, Vinters HV, Tabaton M, Shimohama S, Cash AD, Siedlak SL, Harris PLR, Jones PK, Peterson RB, Perry G, Smith MA (2001) Mitochondrial abnormalities in Alzheimer’s disease. J Neurosci 21:3017-3023.

Hottinger AF, Fine EG, Gurney ME, Zurn AD, Aebischer P (1997) The copper chelator D-penicillamine delays onset of disease and extends survival in a transgenic mouse model of familial amyotrophic lateral sclerosis. Eur. J. Neurosci. 9:1548–1551

Hsu LJ, Sagara Y, Arroyo A, Rockenstein E, Sisk A, Mallory M, Wong J, Takenouchi T, Hashimoto M, Masliah E (2000) alpha-synuclein promotes mitochondrial deficit and oxidative stress. Am J Pathol. 157:401-10.

The Huntington’s Disease Collaborative Research Group (1993) A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosome. Cell 72:971–983.

Huntley GW, Gil O, Bozdagi O (2002) The cadherin family of cell adhesion molecules: multiple roles in synaptic plasticity. Neuroscientist 8:221-233.

Hutton M, Lendon CL, Rizzu P, Baker M, Froelich S, Houlden H, Pickering-Brown S, Chakraverty S, Isaacs A, Grover A, Hackett J, Adamson J, Lincoln S, Dickson D, Davies P, Petersen RC, Stevens M, de Graaff E, Wauters E, van Baren J, et al. (1998) Association of missense and 5'-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 393: 702-705.

Hyun DH, Lee M, Hattori N, Kubo S, Mizuno Y, Halliwell B, Jenner P (2002) Effect of wild-type or mutant Parkin on oxidative damage, nitric oxide, antioxidant defenses, and the proteasome. J Biol Chem. 277:28572-7.

151

Inohara N, Koseki T, Chen S, Wu X, Nunez G (1998) CIDE, a novel family of cell death activators with homology to the 45 kDa subunit of the DNA fragmentation factor. EMBO J 17:2526-2533.

Ishii N, Fujii M, Hartman PS, Tsuda M, Yasuda K, Senoo-Matsuda N, Yanase S, Ayusawa D, Suzuki K (1998) A mutation in succinate dehydrogenase cytochrome b causes oxidative stress and in nematodes. Nature 394:694-697.

Iwatsuji K, Nakamura S, Kameyama M (1989) Lymphocyte glutamate dehydrogenase activity in normal aging and neurological diseases. Gerontology 35:218-224.

Jin P, Zarnescu DC, Zhang F, Pearson CE, Lucchesi JC, Moses K, Warren ST (2003) RNA-mediated neurodegeneration caused by the fragile X premutation rCGG repeats in Drosophila. Neuron 39:739-47.

Kasai H, Chung MH, Jones DS, Inoue H, Ishikawa H, Kamiya H, Ohtsuka E, Nishimura S (1991) 8-Hydroxyguanine, a DNA adduct formed by oxygen radicals: its implication on oxygen radical-involved mutagenesis/. J Toxicol Sci 16 Suppl 1:95-105.

Kato A, Ozawa F, Saitoh Y, Hirai K, Inokuchi K (1997) vesl, a gene encoding VASP/Ena family related protein, is upregulated during seizure, long-term potentiation and synaptogenesis. FEBS Lett 412:183-189.

Keil U, Bonert A, Marques CA, Scherping I, Weyermann J, Strosznajder JB, Muller- Spahn F, Haass C, Czech C, Pradier L, Muller WE, Eckert A (2004) Amyloid beta- induced changes in nitric oxide production and mitochondrial activity lead to apoptosis. J Biol Chem. 279:50310-20.

Keller JN, Mark RJ, Bruce AJ, Blanc E, Rothstein JD, Uchida K, Waeg G, Mattson MP. (1997) 4-Hydroxynonenal, an aldehydic product of membrane lipid peroxidation, impairs glutamate transport and mitochondrial function in synaptosomes. Neuroscience. 80:685-96.

Kikuchi A, Takeda A, Onodera H, Kimpara T, Hisanaga K, Sato N, Nunomura A, Castellani RJ, Perry G, Smith MA, Itoyama Y (2002) Systemic increase of oxidative nucleic acid damage in Parkinson's disease and multiple system atrophy. Neurobiol Dis. 9:244-8.

Kim HC, Yamada K, Nitta A, Olariu A, Tran MH, Mizuno M, Nakajima A, Nagai T, Kamei H, Jhoo WK, Im DH, Shin EJ, Hjelle OP, Ottersen OP, Park SC, Kato K, Mirault ME, Nabeshima T (2003) immunocytochemical evidence that amyloid betat (1-42) impairs endogenous antioxidant systems in vivo. Neuroscience 119:399-419.

152

Kish SJ, Bergeron C, Rajput A, Dozie S, Mastrogiacomo F, Chang LJ, Wilson JM, Distefano LM, Nobrega JN (1992) Cytochrome oxidase in Alzheimer’s disease. J Neurochem. 59:776-779.

Klegeris A, McGeer PL (1994) Rat brain microglia and peritoneal macrophages show similar responses to respiratory burst stimulants. J Neuroimmunol 53:83-90.

Klivenyi P, Andreassen OA, Ferrante RJ, Lancelot E, Reif D, Beal MF (2000) Inhibition of neuronal nitric oxide synthase protects against MPTP toxicity. Neuroreport. 11:1265-8.

Kohno K, Izumi H, Uchiumi T, Ashizuka M, Kuwano M (2003) The pleiotropic functions of the Y-box-binding protein, YB-1. Bioessays 25:691–698.

Koppal T, Drake J, Yatin S, Jordan B, Varadarajan S, Bettenhausen L, Butterfield DA (1999) Peroxynitrite-induced alterations in synaptosomal membrane proteins: insight into oxidative stress in Alzheimer's disease. J Neurochem 72:310-317

Kress GJ, Dineley KE, Reynolds IJ (2002) The relationship between intracellular free iron and cell injury in cultured neurons, astrocytes and oligodendrocytes. J Neurosci. 22:5848-5855.

Liberatore GT, Jackson-Lewis V, Vukosavic S, Mandir AS, Vila M, McAuliffe WG, Dawson VL, Dawson TM, Przedborski S (1999) Inducible nitric oxide synthase stimulates dopaminergic neurodegeneration in the MPTP model of Parkinson disease. Nat Med. 5:1403-9.

Levine RL, William JA, Stadtman ER, Shacter E (1994) Carbonyl assays for determination of oxidatively modified proteins. Methods Enzymol. 233:346-357.

Linnane AW, Marzuki S, Ozawa T, Tanaka M (1989) Mitochonrial DNA mutations as an important contributor to ageing and degenerative diseases. Lancet I:642-645.

Loeffler DA, Connor JR, Juneau PL, Snyder BS, Kanaley L, DeMaggio AJ, Nguyen H, Brickman CM, LeWitt PA (1995) Transferrin and iron in normal, Alzheimer’s disease, and Parkinson’s disease brain regions. J Neurochem 65:710–724.

Lotharius J and Brundin P (2002) Impaired dopamine storage resulting from α- synuclein mutations may contribute to the pathogenesis of Parkinson’s disease. Hum. Mol. Genet. 11:2395–2407.

153

Lovell MA, Ehmann WD, Butler SM, Markesbery WR (1995) Elevated thiobarbituric acid-reactive substances and antioxidant enzyme activity in the brain in Alzheimer's disease. Neurology. 45:1594-601.

Lovell MA, Ehmann WD, Mattson MP, Markesbery WR (1997) Elevated 4- hydroxynonenal in ventricular fluid in Alzheimer's disease. Neurobiol Aging. 18:457- 61.

Lovell MA, Gabbita SP, Markesbery WR (1999) Increased DNA oxidation and decreased levels of repair products in Alzheimer's disease ventricular CSF. J Neurochem. 72:771-6.

Lovell MA, Markesbery WR (2001) Ratio of 8-hydroxyguanine in intact DNA to free 8-hydroxyguanine is increased in Alzheimer disease ventricular cerebrospinal fluid. Arch Neurol. 58:392-6.

Lovell MA, Xie C, Markesbery WR (2000) Decreased base excision repair and increased helicase activity in Alzheimer's disease brain. Brain Res. 855:116-23.

Luo Y, Henle ES, Linn S (1996) Oxidative damage to DNA constituents by iron- mediated fenton reactions. The deoxycytidine family. J Biol Chem. 271:21167-76.

Lyons TJ, Liu H, Goto JJ, Nersissian A, Roe JA, Graden JA, Cafe C, Ellerby LM, Bredesen DE, Gralla EB, Valentine JS (1996) Mutations in copper-zinc superoxide dismutase that cause amyotrophic lateral sclerosis alter the zinc binding site and the redox behavior of the protein. Proc Natl Acad Sci U S A. 93:12240-4.

Markesbery WR (1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Radical Biology and Medicine 23:134-147

Markesbery WR, Lovell MA (1998) Four-hydroxynonenal, a product of lipid peroxidation, is increased in the brain in Alzheimer's disease. Neurobiol Aging. 19:33-6.

Marcus DL, Thomas C, Rodriguez C, Simberkoff K, Tsai JS, Strafaci JA, Freedman ML (1998) Increased peroxidation and reduced antioxidant enzyme activity in Alzheimer's disease. Exp Neurol. 150:40-4.

Mark RJ, Lovell MA, Markesbery WR, Uchida K, Mattson MP (1997) A role for 4- hydroxynonenal, an aldehydic product of lipid peroxidation, in disruption of ion homeostasis and neuronal death induced by amyloid beta-peptide. J Neurochem. 68:255-64.

154

Markesbery WR (1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Radical Biology and Medicine 23:134-147

Mattiazzi M, D'Aurelio M, Gajewski CD, Martushova K, Kiaei M, Beal MF, Manfredi G (2002) Mutated human SOD1 causes dysfunction of oxidative phosphorylation in mitochondria of transgenic mice. J Biol Chem. 277:29626-33.

Mattson MP, Cheng B, Culwell AR, Esch FS, Lieberburg I, Rydel RE (1993) Evidence for excitoprotective and intraneuronal calcium-regulating roles for secreted forms of the beta-amyloid precursor protein. Neuron 10:243-54

McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM (1984) Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 34:939-44.

Mecocci P, Beal MF, Cecchetti R, Polidori MC, Cherubini A, Chionne F, Avellini L, Romano G, Senin U (1997) Mitochondrial membrane fluidity and oxidative damage to mitochondrial DNA in aged and AD human brain. Mol Chem Neuropathol 31:53- 64.

Mecocci P, MacGarvey U, Kaufman AE, Koontz D, Shoffner JM, Wallace DC, Beal MF (1993) Oxidative damage to mitochondrial DNA shows marked age-dependent increases in human brain. Ann Neurol 34:609-616.

Mecocci P, MacGarvey U, Beal MF (1994) Oxidative damage to mitochondrial DNA is increased in Alzheimer's disease. Ann Neurol 36:747-751.

Meda L, Cassatella MA, Szendrei GI, Otvos Jr L, Baron P, Villalba M, Ferrari D, Rossi F (1995) Activation of microglial cells by beta-amyloid protein and interferon- gamma. Nature 374:647-650

Meneghini R (1997) Iron homeostasis, oxidative stress, and DNA damage. Free Radic Biol Med. 23:783-792.

Menzies FM, Ince PG, Shaw PJ (2002) Mitochondrial involvement in amyotrophic lateral sclerosis. Neurochem Int. 40:543-51.

Merry BJ (2002) Molecular mechanisms linking calorie restriction and longevity. International J. Biochem. Cell Biol. 34:1340-1354

Mills J, Laurent Charest D, Lam F, Beyreuther K, Ida N, Pelech SL, Reiner PB (1997) Regulation of amyloid precursor protein catabolism involves the mitogen-activated protein kinase signal transduction pathway. J Neurosci 17:9415-9422.

155

Milzani A, DalleDonne I, Colombo R (1997) Prolonged oxidative stress on actin. Archs Biochem Biophys 339:267-274.

Mirault ME, Nabeshima T (2003) Immunocytochemical evidence that amyloid beta (1-42) impairs endogenous antioxidant systems in vivo. Neuroscience 119:399-419.

Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, Brownlee LM, Vogel FS, Hughes JP, van Belle G, Berg L (1991) The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer's disease. Neurology. 41:479-86.

Montine KS, Quinn JF, Zhang J, Fessel JP, Roberts LJ 2nd, Morrow JD, Montine TJ (2004) Isoprostanes and related products of lipid peroxidation in neurodegenerative diseases. Chem Phys Lipids. 128:117-24.

Montine TJ, Amarnath V, Martin ME, Strittmatter WJ, Graham DG (1996) E-4- hydroxy-2-nonenal is cytotoxic and cross-links cytoskeletal proteins in P19 neuroglial cultures. Am J Pathol. 148:89-93.

Moore D, Iritani S, Chambers J, Emson P (2000) Immunohistochemical localization of the P2Y1 purinergic receptor in Alzheimer's disease. Neuroreport 11:3799-3803.

Mullart E, Boerrigter METI, Ravid R, Swaab DF, Vijg J (1990) Increased levels of DNA breaks in cerebral cortex of Alzheimer’s disease patients. Neurobiol. Aging 11:169-173.

Multhaup G, Schlicksupp A, Hesse L, Beher D, Ruppert T, Masters CL, Beyreuther K (1996) The amyloid precursor protein of Alzheimer’s disease in the reduction of Copper (II) to Copper (I). Science 271:1406-1409.

Namekata K, Imagawa M, Terashi A, Ohta S, Oyama F, Ihara Y (1997) Association of transferrin C2 allele with late-onset Alzheimer’s disease. Hum Genet 101:126–129.

Nunomura A, Perry G, Pappolla MA, Wade R, Hirai K, Chiba S, Smith MA (1999) RNA oxidation is a prominent feature of vulnerable neurons in Alzheimer's disease. J Neurosci 19:1959-1964.

Nixon RA (2000) A "protease activation cascade" in the pathogenesis of Alzheimer's disease. Ann N Y Acad Sci 924:117-131.

Ohta S, Mineta T, Kimoto M, Tabuchi K (1997) Differential Display Cloning of a Novel Rat cDNA (RNB6) That Shows High Expression in the Neonatal Brain Revealed a Member of Ena/VASP Family. Biochem Biophys Res Commun 237:307- 312.

156

Omar RA (1999) Increased Expression but Reduced Activity of Antioxidant Enzymes in Alzheimer's disease. J Alzheimers Dis 1:139-145.

Ougland R, Zhang CM, Liiv A, Johansen RF, Seeberg E, Hou YM, Remme J, Falnes PO (2004) AlkB restores the biological function of mRNA and tRNA inactivated by chemical methylation. Mol Cell. 2004 16:107-16.

Panov AV, Burke JR, Strittmatter WJ, Greenamyre JT (2003) In vitro effects of polyglutamine tracts on Ca2+-dependent depolarization of rat and human mitochondria: Relevance to Huntington’s disease. Arch. Biochem. Biophys. 410:1–6.

Paragh G, Balla P, Katona E, Seres I, Egerhazi A, Degrell I. (2002) Serum paraoxonase activity changes in patients with Alzheimer's disease and vascular dementia. Eur Arch Psychiatry Clin Neurosci 252:63-67.

Parker WD Jr, Parks J, Filley CM, Kleinschmidt-DeMasters BK (1994) Electron transport chain defects in Alzheimer's disease brain. Neurology. 44:1090-6.

Parks JK, Smith TS, Trimmer PA, Bennett JP Jr, Parker WD Jr. (2001) Neurotoxic Abeta peptides increase oxidative stress in vivo through NMDA-receptor and nitric- oxide-synthase mechanisms, and inhibit complex IV activity and induce a mitochondrial permeability transition in vitro. J Neurochem. 2001 76:1050-6.

Partridge RS, Monroe SM, Parks JK, Johnson K, Johnson WD, Johnson Jr WD, Eaton GR, Eaton SS (1994) Spin trapping of azidyl and hydroxyl radicals in azide- inhibited rat brain submitochondrial particles. Arch. Biochem. Biophys 310:210-217.

Perry TL and Yong VW (1986) Idiopathic Parkinson's disease, progressive supranuclear palsy and glutathione metabolism in the substantia nigra of patients. Neurosci. Lett. 67, 269–274.

Petersen RC, Stevens JC, Ganguli M, Tangalos EG, Cummings GL, Dekosky ST (2001) Practice parameters: early detection of dementia in mild cognitive impairment (an evidence based review), Neurology 56:1133–1142.

Picklo MJ, Amarnath V, McIntyre JO, Graham DG, Montine TJ (1999) 4-Hydroxy- 2(E)-nonenal inhibits CNS mitochondrial respiration at multiple sites. J Neurochem. 72:1617-24.

Poirier J (1994) Apolipoprotein E in animal models of CNS injury and in Alzheimer's disease. Trends Neurosci. 17:525-30.

157

Polidori MC, Mecocci P, Browne SE, Senin U, and Beal MF (1999) Oxidative damage to mitochondrial DNA in Huntington’s disease parietal cortex. Neurosci. Lett. 272:53–56.

Pratico D and Sung Syun (2004) Lipid peroxidation and oxidative imbalance: early functional events in Alzheimer’s disease. J Alzheimer’s Disease 6:171-175.

Price DL, Tanzi RE, Borchelt DR, Sisodia SS (1998) Alzheimer's disease: genetic studies and transgenic models. Annu Rev Genet. 32:461-93.

Ramassamy C, Krzywokowski P, Bastianetteo S, et al. (1998) Apolipoprotein E, oxidative stress and EGb761 in Alzheimer’s disease brain. In: Packer L, Christen Yeds. Ginkgo biloba extract (EGb761) study: lesson from . Paris: Elsevier 69-83

Ranum LP and Day JW (2002) Dominantly inherited, non-coding microsatellite expansion disorders. Curr. Opin. Genet. Dev. 12:266-271.

Reich EE, Markesbery WR, Roberts LJ 2nd, Swift LL, Morrow JD, Montine TJ (2001) Brain regional quantification of F-ring and D-/E-ring isoprostanes and neuroprostanes in Alzheimer's disease. Am J Pathol. 2001 Jan;158:293-7.

Rhee Y, Valentine MR, Termini J (1995) Oxidative base damage in RNA detected by reverse transcriptase. Nucleic Acids Res 23:3275-3282.

Roder HM, Eden PA, Ingram VM (1993) Brain protein kinase PK 40 converts TAU into a PHF-like form as found in Alzheimer’s disease. Biochem Biophys Res Commun 193:639–647.

Rosen DR, Siddique T, Patterson D, Figlewicz DA, Sapp P, Hentati A, Donaldson D, Goto J, O’Regan JP, Deng HX, Rahmani Z, Krizus A, McKenna-Yasek D, Cayabyas A, Gaston SM, Berger R, Tanzi RE, Halperin JJ, Herzfeldt B, Bergh RVN, Hung WY, Bird T, Deng G, Mulder DW, Smyth C, Laing NG, Soriano E, Pericak-Vance MA, Haines J, Rouleau GA, Gusella JS, Horvitz HR, Brown RH (1993) Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362:59-62.

Sadot E, Jaaro H, Seger R, Ginzburg I (1998) Ras-signaling pathways: positive and negative regulation of tau expression in PC12 cells. J Neurochem 70:428-431.

Saunders RD, Luttman CA, Keith PT, Little SP (1991) Beta-amyloid protein potentiates H2O2-induced neuron degeneration in vitro. Soc Neurosci. Abstr. 17:1447

158

Sayre LM, Zelasko DA, Harris PL, Perry G, Salomon RG, Smith MA (1997) 4- Hydroxynonenal-derived advanced lipid peroxidation end products are increased in Alzheimer's disease. J Neurochem 68:2092-2097.

Shan X, Tashiro H, Lin CL (2003) The identification and characterization of oxidized RNAs in Alzheimer's disease. J Neurosci 23:4913-21.

Siems WG, Hapner SJ, van Kuijk FJ (1996) 4-hydroxynonenal inhibits Na(+)-K(+)- ATPase. Free Radic Biol Med. 20:215-23.

Sies H (1991) Oxidative stress: oxidants and antioxidants. Academin Press, London. Sun F, Shan X, Bresnahan JC, Lin CL, Beattie MS (2003) RNA oxidation after rat spinal cord cotusion injury. Soc. Neurosci. Abs. 953.13.

Smith CD, Carney JM, Starke-Reed PE, Oliver CN, Stadtman ER, Floyd RA, Markesbery WR (1991) Excess brain protein oxidation and enzyme dysfunction in normal aging and in Alzheimer disease. Proc Natl Acad Sci U S A 88:10540-10543.

Smith MA, Harris PL, Sayre LM, Perry G (1997) Iron accumulation in Alzheimer disease is a source of redox-generated free radicals. Proc Natl Acad Sci U S A 94:9866-9868.

Smith MA, Richey Harris PL, Sayre LM, Beckman JS, Perry G (1997) Widespread peroxynitrite-mediated damage in Alzheimer's disease. J Neurosci 17:2653-2657

Smythies J (1996) On the function of neuromelanin. Proc. R. Soc. Lond. B Biol. Sci. 263:487–489.

Sofic E, Paulus W, Jellinger K, Riederer P, Youdim MB (1991) Selective increase of iron in substantia nigra zona compacta of parkinsonian brains. J. Neurochem. 56:978– 982.

Spillantini MG, Schmidt ML, Lee VM, Trojanowski JQ, Jakes R, Goedert M (1997) α-synuclein in Lewy bodies. Nature 388:839–840.

Stadtman ER (1990) Metal ion-catalyzed oxidation of proteins: biochemical mechanism and biological consequences. Free Radic Biol Med. 9:315-25.

Subbarao KV, Richardson JS, Ang LC (1990) Autopsy samples of Alzheimer's cortex show increased peroxidation in vitro. J Neurochem 55:342-345.

Sulzer D, Bogulavsky J, Larsen KE, Behr G, Karatekin E, Kleinman MH, Turro N, Krantz D, Edwards RH, Greene LA, Zecca L (2000) Neuromelanin biosynthesis is

159

driven by excess cytosolic catecholamines not accumulated by synaptic vesicles. Proc. Natl Acad. Sci. USA 97:11869–11874.

Tanaka M (2002) Mitochondrial genotypes and cytochrome b variants associated with longevity or Parkinson's disease. J Neurol 249 suppl 2:II11-18.

Tanaka M, Sotomatsu A, Yoshida T, Hirai S, Nishida A (1994) Detection of superoxide production by activated microglia using a sensitive and specific chemiluminescence assay and microglia-mediated PC12h cell death. J Neurochem 63:266-270

Terrisse L, Poirier J, Bertrand P, Merched A, Visvikis S, Siest G, Milne R, Rassart E (1998) Increased levels of apolipoprotein D in cerebrospinal fluid and hippocampus of Alzheimer's patients. J Neurochem 71:1643-1650.

Thomas T, Thomas G, McLendon C, Sutton T, Mullan M (1996) β-amyloid-mediated vasoactivity and vascular endothelial damage. Nature 380:168-171.

Tipton KF and Singer TP (1993) Advances in our understanding of the mechanisms of the neurotoxicity of MPTP and related compounds. J. Neurochem. 61:1191-1206.

Tohgi H, Abe T, Yamazaki K, Murata T, Ishizaki E, Isobe C (1999) Alterations of 3- nitrotyrosine concentration in the cerebrospinal fluid during aging and in patients with Alzheimer’s disease. Neurosci. Lett. 269:52-54.

Tritschler HJ, Medori R (1993) Mitochondrial DNA alterations as a source of human disorders. Neurology 43:280-288.

Troncoso JC, Costello A, Watson AL Jr, Johnson GV (1993) In vitro polymerization of oxidized tau into filaments. Brain Res. 613:313-316. van Rensburg SJ, Potocnik FC, De Villiers JN, Kotze MJ, Taljaard JJ (2000) Earlier age of onset of Alzheimer’s disease in patients with both the transferring C2 and apolipoprotein E-epsilon 4 alleles. Ann NY Acad Sci 903:200–203.

Vassar R, Bennett BD, Babu-Khan S, Kahn S, Mendiaz EA, Denis P, Teplow DB, Ross S, Amarante P, Loeloff R, Luo Y, Fisher S, Fuller J, Edenson S, Lile J, Jarosinski MA, Biere AL, Curran E, Burgess T, Louis JC, Collins F, Treanor J, Rogers G, Citron M (1999) Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science. 286:735-41. von Eitzen U, Kosel S, Grasbon-Frodl EM, Egensperger, R, Graeber MB (2000) Sequence analysis of the MTCYB gene in Parkinson disease. Neurogenetics 3:47-48.

160

Wamer WG, Yin JJ, Wei RR (1997) Oxidative damage to nucleic acids photosensitized by titanium dioxide. Free Radic Biol Med. 23:851-858.

Wiedau-Pazos M, Goto JJ, Rabizadeh S, Gralla EB, Roe JA, Lee MK, Valentine JS, Bredesen DE (1996) Altered reactivity of superoxide dismutase in familial amyotrophic lateral sclerosis. Science. 271:515-8.

Wood PL (1995) Microglia as a unique cellular target in the treatment of stroke: potential neurotoxic mediators produced by activated microglia. Neurol Res 17:242- 248

Xie Z, Wei M, Morgan TE, Fabrizio P, Han D, Finch CE, Longo VD (2002) Peroxynitrite mediates neurotoxicity of amyloid beta-peptide1-42- and lipopolysaccharide-activated microglia. J Neurosci. 22:3484-92.

Yamada K, Tanaka T, Han D, Senzaki K, Kameyama T, Nabeshima T Protective effects of idebenone and alpha-tocopherol on beta-amyloid-(1-42)-induced learning and memory deficits in rats: implication of oxidative stress in beta-amyloid-induced neurotoxicity in vivo. Eur J Neurosci. 11:83-90.

Yamakura F, Taka H, Fujimura T, Murayama K (1998) Inactivation of human manganese-superoxide dismutase by peroxynitrite is caused by exclusive nitration of tyrosine 34 to 3-nitrotyrosine. J Biol Chem. 273:14085-9.

Yim MB, Kang JH, Yim HS, Kwak HS, Chock PB, Stadtman ER (1996) A gain-of- function of an amyotrophic lateral sclerosis-associated Cu,Zn-superoxide dismutase mutant: An enhancement of free radical formation due to a decrease in Km for hydrogen peroxide. Proc Natl Acad Sci U S A. 93:5709-14.

Yu TW, Anderson D (1997) Reactive oxygen species-induced DNA damage and its modification: a chemical investigation. Mutat Res. 379:201-10.

Zarkovic K (2003) 4-hydroxynonenal and neurodegenerative diseases. Mol Aspects Med. 24:293-303.

Zhang J, Perry G, Smith MA, Robertson D, Olson SJ, Graham DG, Montine TJ (1999) Parkinson’s disease is associated with oxidative damage to cytoplasmic DNA and RNA in substantia nigra neurons. Am J Pathol. 154:1423-9.

161