The Pennsylvania State University
The Graduate School
College of Medicine
MODELING THE IMPACT OF H63D HFE POLYMORPHISM ON
AMYOTROPHIC LATERAL SCLEROSIS (ALS)
A Dissertation in
Neuroscience
by
Wint Nandar
© 2013 Wint Nandar
Submitted in Partial Fulfillment
of the Requirements
for the Degree of
Doctor of Philosophy
August 2013 ii
The dissertation of Wint Nandar was reviewed and approved* by the following:
James R. Connor University Distinguished Professor Vice Chair of Department of Neurosurgery Dissertation Advisor Chair of Committee
Zachary Simmons Professor of Neurology
Robert H. Bonneau Professor of Microbiology and Immunology
Sang Y. Lee Assistant Professor of Neurosurgery Special Member
Patricia S. Grigson Professor of Neural and Behavioral Sciences (Committee member and Director of Neuroscience Graduate Program)
*Signatures are on file in the Graduate School. iii
ABSTRACT
Amyotrophic lateral sclerosis (ALS), commonly known as Lou Gehrig’s disease, is a progressive neurodegenerative disorder characterized by selective degeneration of upper and lower motor neurons in the brain and spinal cord. Degeneration of motor neurons results in progressive paralysis eventually causing the inability to walk, talk, swallow and breathe. The average survival is 2-5 years after disease onset mostly due to respiratory failure. The etiology underlying or the genetic basis for the pathogenesis of sporadic ALS (sALS), which accounts for the majority of all cases, remains unclear. In
2004 our laboratory demonstrated that the H63D polymorphism in the HFE gene is found in as many as 30% of ALS patients. After a number of other laboratories performed similar studies, a meta-analysis indicates that the presence of the H63D HFE polymorphism increases the risk of ALS 4-fold. Therefore, the main objective of this thesis is to evaluate the effect of H63D HFE on ALS pathogenesis. We generated two animal models: a mouse model carrying H67D HFE (homologous to human H63D) and a double mutant mouse line (SOD1/H67D) carrying both H67D HFE and SOD1(G93A) mutations. The latter mutation is well-established as inducing the ALS phenotype in mice. The current paradigm relating to HFE gene variants holds that the blood-brain- barrier protects the brain from loss of iron homeostasis resulting from HFE mutations.
Therefore, our in vivo models provide opportunities to challenge this paradigm and establish new models for neuroscience research.
Normally HFE protein interacts with transferrin receptor (TfR) on the cell surface and modulates iron uptake. The H63D HFE gene variant, a frequently found allelic iv
variant in Caucasians, fails to limit TfR-mediated iron uptake. Although increasing evidence suggest that H63D HFE is associated with neurodegenerative diseases, how
H63D HFE impacts neuronal degeneration is not fully understood. Therefore, we generated a H67D knock-in mouse line (homologous to human H63D) using site-directed mutagenesis of the murine HFE gene to evaluate the neurological consequences of H63D
HFE in vivo. We hypothesized that the HFE gene variant alters brain iron homeostasis and creates an environment for increased oxidative stress. Although total brain iron concentration does not change significantly in H67D mice, brain iron management proteins expression is significantly altered. For example, the iron storage protein ferritin is increased indicating an attempt by the brain to sequester iron to protect the brain from iron induced oxidative stress. Transferrin, the iron mobilization protein is decreased which is consistent with increased iron in the brain. A subunit of ferritin (L-subunit) is increased in microglia in H67D mice suggesting these cells are activated and attempting to contribute to brain iron homeostasis by storing excess iron. However, at younger ages, these compensatory responses are not sufficient to protect the brain as reflected by increased gliosis and oxidative stress in 6-month-old mice. Absence of increased oxidative stress at 12 months may be associated with adaptive responses mediated by nuclear factor E2-related factor 2 (Nrf2), which regulates antioxidant genes expression and is increased in the brains of H67D mice. Our findings indicate that, in contrast to the classical paradigm, H63D HFE alters brain iron metabolism and promotes an environment for oxidative stress. The H67D knock-in mice will become a preclinical model to evaluate the role of H63D HFE in neurodegenerative diseases including ALS, which we analyzed in the second part of this study. v
The H63D HFE variant is present in as many as 30% of ALS patients. To test our hypothesis that H63D HFE modifies ALS disease pathogenesis we generated a double mutant mouse line (SOD1/H67D) carrying both H67D HFE and SOD1(G93A) mutations.
The SOD1/H67D mice have a shorter survival and an accelerated disease progression compared to SOD1 mutant mouse model of ALS (G93A). To determine mechanisms underlying the accelerated disease, I measured expression of proteins involved in iron homeostasis and oxidative stress in lumbar spinal cords of SOD1/H67D mice at 90 days
(presymptomatic), 110 days (symptomatic) and end-stage. Expression of TfR, a cellular iron uptake protein, is decreased in both SOD1/H67D and SOD1 mice starting at 90 days.
However, in SOD1/H67D mice decreased TfR is accompanied by increased L-ferritin levels suggesting altered iron homeostasis in these mice. The increase in L-ferritin coupled with an increase in caspase-3 levels are suggestive of microglial activation in these mice. Although increased oxidative stress and gliosis are present in both groups, the magnitude of the increase is higher in the SOD1/H67D compared with SOD1 mice, particularly during the symptomatic stage. Our findings suggest that elevated oxidative stress and increased gliosis associated with altered iron homeostasis are underlying mechanisms contributing to an accelerated disease in SOD1/H67D mice. These results support the argument that H63D HFE is a contributing factor in ALS pathogenesis.
Together, our studies strongly suggest that H63D HFE is a genetic modifier for
ALS independent of environmental factors and causes functional consequences in the brain and the spinal cord. These animal models can be used as a preclinical model to study how gene-environment interactions impact disease mechanisms and will aid in vi
developing treatment strategies incorporating patient stratification by HFE genotype when assessing therapeutic interventions.
vii
TABLE OF CONTENTS
LIST OF TABLES……………………………………………………………………...... x
LIST OF FIGURES………………………………………………………………………xi
LIST OF ABBREVIATIONS…………………………………………………………..xiii
GLOSSARY OF THE TERMS………………………………………………………….xv
ACKNOWLEDGEMENTS……………………………………………………………..xvi
CHAPTER 1. HFE Gene Variants Affect Iron in the Brain………………………...... 1
1.1. Abstract……………………………………………………………………….1 1.2. Introduction………………………………………………………………...... 2 1.3. Iron in the Brain………………………………………………………………3 1.4. The HFE Gene………………………………………………………………..7 1.5. The Structure of HFE Protein………………………………………………...8 1.6. Function of HFE Protein……………………………………………………...8 1.7. HFE and Neurodegenerative Disorders……………………………………..11 1.7.1. HFE and Amyotrophic Lateral Sclerosis………………………...12 1.7.2. HFE and Alzheimer’s Disease…………………………………...16 1.7.3. HFE and Parkinson’s Disease……………………………………24 1.7.4. HFE and Ischemic Stroke………………………………………..29 1.8. Animal Models………………………………………………………………33 1.9. References…………………………………………………………………...35
CHAPTER 2. H63D HFE gene variant alters brain iron profiles and increases oxidative stress ………………………………………………55
2.1. Abstract……………………………………………………………………...55 2.2. Introduction………………………………………………………………….56 2.3. Materials and Methods………………………………………………………58 2.3.1. Generation of H67D Knock-in Mice……………………………...58 2.3.2. Mice Genotyping………………………………………………….59 2.3.3. Measurement of Iron………………………………………………59 2.3.4. Immunoblotting……………………………………………………59 2.3.5. Histology…………………………………………………………..60 2.3.6. Myelin Isolation and Analysis of Myelin Proteins………………..62 2.3.7. Measurement of Oxidatively Modified Proteins…………………..63 2.3.8. Statistical Analyses………………………………………………..63 viii
2.4. Results……………………………………………………………………….63 2.4.1. Generation of H67D mice…………………………………………63 2.4.2. Brain Iron Concentration in H67D mice…………………………..70 2.4.3. Alteration in Expression of Proteins Invovled in Iron Homeostasis in H67D Mice…………………………………..72 2.4.4. Myelin Proteins Were Not Changed in H67D Mice………………96 2.4.5. H67D HFE is Associated with Astrogliosis……………………….99 2.4.6. H67D HFE is Associated with Oxidative Stress…………………104 2.5. Discussion………………………………………………………………….110 2.6. Acknowledgements………………………………………………………...115 2.7. References………………………………………………………………….116
Chapter 3 H63D HFE accelerates disease progression in Animal Models of amyotrophic lateral sclerosis………………………………………………124
3.1. Abstract…………………………………………………………………….124 3.2. Introduction………………………………………………………………...125 3.3. Material and Methods……………………………………………………...127 3.3.1. Animal Model: Double Mutant SOD1/H67D Mice Generation………………………………………………………..127 3.3.2. HFE and SOD1(G93A) Genotyping……………………………..128 3.3.3. Behavior and Survival……………………………………………129 3.3.3.1. Rotarod…………………………………………………129 3.3.3.2. Grip Strength………………………………………...... 129 3.3.3.3. Survival and Disease Duration…………………………130 3.3.4. Histology…………………………………………………………130 3.3.5. Western Blot……………………………………………………..131 3.3.6. Statistical Analyses………………………………………………132 3.4. Results……………………………………………………………………...132 3.4.1. H67D HFE Shortens Survival and Accelerates Disease Progression………………………………………………132 3.4.2. H67D HFE has no Effect on Motor Neuron Loss………………..144 3.4.3. Altered Iron Management Protein Expression in Double Mutant (SOD1/H67D) Mice………………..……...... 146 3.4.4. Increased Oxidative Stress in Double Mutant (SOD1/H67D) Mice……………………………………………...151 3.4.5. Increased Caspase-3 in Double Mutant (SOD1/H67D) Mice……………………………………………...154 3.4.6. Increased Gliosis in Double Mutant (SOD1/H67D) Mice……………………………………………...156 3.5. Discussion………………………………………………………………….159 3.6. Acknowledgements………………………………………………………...165 3.7. References………………………………………………………………….166 ix
Chapter 4 H63D HFE – a genetic modifier for risk of ALS: summary and future perspectives…………………………………………176
4.1. Introduction and Summary………………………………………………...176 4.2. Activating the Nrf2 Signaling Pathway as a Potential Therapy for ALS….179 4.3. Caspase Inhibitors Hold Promise for Therapeutic Benefits in ALS……….185 4.4. Hepcidin may be Neuroprotective in ALS…………………………………187 4.5. Applications to the Human Population………………..…………………...190 4.6. Conclusions………………………………………………………………...194 4.7. References………………………………………………………………….196
x
LIST OF TABLES
Table 1-1 Neurodegeneration with brain iron accumulation caused by mutations in genes involved in iron metabolism…………………………6
Table 1-2 Studies evaluating an association between HFE mutations and amyotrophic lateral sclerosis…………………………………………….14
Table 1-3 Studies showing an association between HFE mutations and Alzheimer’s disease……………………………………………………..20
Table 1-4 Studies showing no association between HFE mutations and Alzheimer’s disease……………………………………………………..23
Table 1-5 Studies showing an association between the C282Y HFE variant and Parkinson’s disease……………………………………………………...26
Table 1-6 Studies showing no association of H63D and C282Y HFE with Parkinson’s disease……………………………………………………..28
Table 1-7 Studies evaluating an association between HFE mutations and ischemic stroke…………………………………………………………..32
Table 2-1 Summary findings from neurological characterization of the H67D mice……………………………………………………………...109
Table 3-1 Summary of findings demonstrating pathogenic pathways underlying accelerated disease in double mutant mice…………………………………158 xi
LIST OF FIGURES
Figure 1-1. Schematic diagram for cellular iron metabolism in the presence of wild-type (WT) HFE, H63D HFE and C282Y HFE………………………..10
Figure 2-1. Increased body weight and hepatic iron concentration in H67D knock-in mice…………………………………………………….65
Figure 2-2. Altered hepatic iron management protein expression in H67D knock-in mice…………………………………………………….68
Figure 2-3. Total brain iron concentration is altered in H67D knock-in mice……………………………………………………….71
Figure 2-4. Altered brain iron management protein expression in 6-month-old H67D knock-in mice………………………………………73
Figure 2-5. Altered brain iron management protein expression in 12-month-old H67D knock-in mice……………………………………..76
Figure 2-6. Immunohistochemical localization for transferrin receptor (TfR)…………79
Figure 2-7. Immunohistochemical localization for divalent metal transporter-1 (DMT-1)…………………………………………………………………….82
Figure 2-8. Immunohistochemical localization for transferrin…………………………85
Figure 2-9. Immunohistochemical localization for L-ferritin…………………………..88
Figure 2-10. Double immunostaining showing the co-localization of L-ferritin and Iba-1…………………………………………………………………..91
Figure 2-11. Immunofluorescence staining for Iba-1…………………………………...93
Figure 2-12. Double immunostaining showing no co-localization of L-ferritin and GFAP…………………………………………………………………94
Figure 2-13. Myelin proteins are not altered in H67D knock-in mice…………………..97
Figure 2-14. Increased GFAP in 6-month-old H67D knock-in mice………………….100
Figure 2-15. GFAP levels are not increased in 12-month-old H67D knock-in mice……………………………………………………..102
xii
Figure 2-16. Increased oxidative stress in H67D knock-in mice………………………105
Figure 2-17. H67D HFE increases Nrf2 expression…………………………………...108
Figure 3-1. A representative gel for genotyping of a double mutant mouse line (SOD1/H67D)…………………………………………………133
Figure 3-2. H67D HFE shortens survival in double mutant (SOD1/H67D) mice………………………………………………………135
Figure 3-3. Shorter disease duration in double mutant (SOD1/H67D) mice……...... 138
Figure 3-4. Age of disease onset is not altered in double mutant (SOD1/H67D) mice……………………………………………………….140
Figure 3-5. Accelerated disease progression in double mutant (SOD1/H67D) mice………………………………………………………142
Figure 3-6. Motor neuron loss in SOD1(G93A) and SOD1/H67D mice……………...145
Figure 3-7. Decreased transferrin receptor (TfR) expression in lumbar spinal cords of double mutant (SOD1/H67D) mice……………………....147
Figure 3-8. Increased L-ferritin expression in lumbar spinal cords of double mutant (SOD1/H67D) mice……………………………………....149
Figure 3-9. H-ferritin expression is not changed in lumbar spinal cords of double mutant (SOD1/H67D) mice…………………………………....150
Figure 3-10. Increased oxidative stress in lumbar spinal cords of double mutant (SOD1/H67D) mice……………………………………..152
Figure 3-11. Increased caspase-3 expression in lumbar spinal cords of double mutant (SOD1/H67D) mice……………………………………..155
Figure 3-12. Increased GFAP expression in lumbar spinal cords of double mutant (SOD1/H67D) mice……………………………………..157
Figure 4-1. Cellular consequences of H63D HFE…………………………………….178
Figure 4-2. Schematic diagram of the Nrf2 signaling pathway……………………….184
Figure 4-3. Schematic diagram of future perspectives………………………………..195 xiii
LIST OF ABBREVIATIONS
ABC Avidin biotin complex AD Alzheimer’s disease ALS amyotrophic lateral sclerosis ANOVA analysis of variance APP amyloid precursor protein APOE apolipoprotein E ARE antioxidant response element BBB Blood brain barrier BCA bicinchoninic acid BMP bone morphogenetic protein CNS central nervous system C9ORF72 chromosome 9 open reading frame 72 DAB 3, 3′-diaminobenzidine DFO desferroxamine DMT-1 divalent metal transporter 1 DNP dinitrophenyl DWI diffusion weighted imaging ECL enhanced chemiluminescent ER endoplasmic reticulum ES embryonic stem cell line FAD familial form of Alzheimer’s disease FALS familial form of amyotrophic lateral sclerosis FUS/TLS fused in sarcoma/translated in liposarcoma GFAP glial fibrillary acidic protein GSH glutathione GSK glycogen synthase kinase HH hereditary hemochromatosis HLA human leukocyte antigen HO-1 hemeoxygenase-1 HPC nondemented controls with AD-like pathology HRP horseradish peroxidase iNOS nitric oxide synthase IL interleukin IRP iron regulatory protein Keap 1 Kelch-like ECH-associated protein 1 LPS lipopolysaccharide MBP myelin basic protein MCI mild cognitive impairment MCP-1 monocyte chemoattractant protein-1 MHC-1 major histocompatibility complex class-1 MRI magnetic resonance imaging NFT neurofibrillary tangle xiv
Nrf2 nuclear factor E2-related factor 2 •OH hydroxyl radical PCR polymerase chain reaction PLP proteolipid protein Pin1 prolyl-peptidyl isomerase 1 ROS reactive oxygen species SAD sporadic form of Alzheimer’s disease SALS sporadic form of amyotrophic lateral sclerosis SE standard error SEM standard error mean siRNA small interfering RNA SOD superoxide dismutase TDP 43 TAR-DNA-binding protein 43 Tf transferrin TfR transferrin receptor Tim-2 T cell immunoglobulin and mucin domain-containing protein-2 TK thymidine kinase WT wild-type xCT cystine/glutamate anitporter.
xv
GLOSSARY OF TERMS*
Allele One of the variant forms of a gene
Genotype Genetic information of a cell or an organism
Knock-in Targeted insertion of a gene
Mutant An organism or a cell carrying a mutation
Polymorphism Genotypic variation within a population
Transgenic Random integration of a gene into the germ line of an animal
*Adapted from The Jackson Laboratory xvi
ACKNOWLEDGEMENTS
I dedicate this thesis to my dearest respectful father, U Kyaw Win Aung, who always believes in me, and supports me while I am pursuing the dream of becoming a scientist away from home. I prepare this thesis in memory of my beloved mother,
Daw Nu Nu Swe – your love and teaching are always in my heart and are forever cherished. This thesis is also dedicated to my brothers, Pyae Phyo Hein and Naing Ko Ko
Lin, who always protects and supports me in their own ways – thank you for standing by me during hardships. Finally, I dedicate this thesis to my husband, Xiaowei William Su who is always next to me sharing my happy and sad moments, being patient with me and understanding me while putting up with my schedule – your support and encouragement is forever appreciated.
I wish to thank my advisor, Dr. James Connor, who is a great teacher, advisor as well as a great mentor. Dr. Connor – you have been very supportive and understanding throughout these years. Under your teaching and guidance, I have made my way up to this point. To my present and past committee members – thank you all for your advice and guidance throughout these years. They are very valuable and help me improve my knowledge.
I would like to express my sincere gratitude towards the past and present members of Connor’s lab. You all have made my everyday lively and thank you all for your patience and your kindness. Particularly, I wish to sincerely thank Beth, Stephanie,
Mandy, Padma and Fatima, who are like my “family” in Hershey – your friendship and accompany are always appreciated. xvii
I would like to take this opportunity to thank Beth Neely for helping me with behavior studies and Dr. Eric Unger for measuring total iron level using atomic absorption spectrophotometry for me at University Park. 1
Chapter 1
HFE Gene Variants Affect Iron in the Brain
1.1 Abstract
Iron accumulation in the brain and increased oxidative stress are consistent observations in many neurodegenerative diseases. Thus, we have begun examination into gene mutations or allelic variants expressed in people with neurological diseases that could be associated with loss of brain iron homeostasis. One of the mechanisms leading to iron overload is a mutation in the HFE gene, which is involved in iron metabolism.
The two most common HFE gene variants are C282Y (1.9 %) and H63D (8.1 %). The
C282Y HFE variant is more commonly associated with hereditary hemochromatosis, which is an autosomal recessive disorder, characterized by iron overload in a number of systemic organs. The H63D HFE variant appears less frequently associated with hemochromatosis, but its role in the neurodegenerative diseases has received more attention. At the cellular level, the HFE mutant protein resulting from H63D HFE gene variant is associated with iron dyshomeostasis, increased oxidative stress, glutamate release, tau phosphorylation, alteration in inflammatory response and prolonged ER stress, each of which is under investigation as a contributing factor to neurodegenerative diseases. Therefore, the HFE gene variants are proposed to be genetic modifiers or risk factors for neurodegenerative diseases by establishing a permissive milieu for pathogenic agents. This review will discuss the current knowledge of the association of the HFE gene variants with neurodegenerative diseases: amyotrophic lateral sclerosis (ALS),
Alzheimer’s disease (AD), Parkinson’s disease (PD) and ischemic stroke. 2
1.2. Introduction
Iron plays a significant role in many biological functions essential for life. The citric acid cycle and electron transport chain of mitochondria contains several iron- dependent enzymes and complexes such as cytochromes, succinate dehydrogenase,
NADH-dehydrogenase, and aconitase. The activity of ribonucleotide reductase, which catalyzes the essential step of DNA synthesis, is dependent on iron [1]. Iron is an indispensable component for neurotransmitter synthesis and myelinogenesis [1, 2].
Because iron is a required-cofactor in cholesterol synthesis, a key component of myelin, and is essential for oxidative metabolic activity of oligodendrocytes that produce myelin
[3, 4], iron deficiency affects myelin production and composition in white matter. In addition to brain morphology, iron deficiency causes an alteration in norepinephrine and dopamine metabolism, which affects neurochemistry and may delay central nervous system development [5].
Though iron is an essential cofactor for many proteins in the central nervous system, free or unbound iron can serves as a pro-oxidant. Ferrous iron (Fe2+) catalyzes the conversion of reactive oxygen species (ROS) to highly reactive hydroxyl radical
(•OH) via Fenton reaction while ferric iron (Fe3+) can react with superoxide (O2•-) and generates Fe2+ leading to •OH formation via the Heber-Weiss reaction. Excess iron can cause protein peroxidation, lipid peroxidation and DNA oxidation, which eventually can lead to cellular and neuronal damage or death [6, 7]. Therefore, iron content in the body and in the CNS is strictly regulated via expression of several proteins involved in its transport, storage and utilization [8, 9]. 3
One iron regulatory protein that is receiving increased attention in neuroscience is the HFE protein. Mutations in the HFE gene are commonly associated with the iron overload genetic disorder hereditary hemochromatosis [10, 11]. Because of their association with iron accumulation, the HFE gene mutations are being investigated as genetic risks for neurodegenerative disorders [6, 9, 11-13]. In this review, we will discuss iron physiology in the brain in relation to HFE structure and function and describe the relationship between HFE gene variants and four neurodegenerative disorders; amyotrophic lateral sclerosis (ALS), Alzheimer’s disease (AD), Parkinson’s disease
(PD), and ischemic stroke.
1.3. Iron in the Brain
The iron concentration in the brain is second to the liver among the organs in the body [14]. Iron is distributed heterogeneously throughout the brain with highest concentration found in the globus pallidus followed by the red nucleus, substantia nigra, putamen and the dentate nucleus of the cerebellum [14, 15]. In the cortical regions, the highest iron concentration was found in the motor cortex followed by the occipital cortex, sensory cortex and the parietal cortex [14]. Brain iron content increases with advancing age [2, 14, 16] and iron content rapidly increases in all brain regions except medulla oblongata during the first two decades of life [14]. It is noteworthy that functionally, the regions of the brain that have the highest iron content are all involved in motor functions.
In addition to its high iron content, the brain generates a large amount of reactive oxygen species (ROS) as a consequence of its high oxygen consumption (20% of body’s total resting oxygen consumption) to meet its high metabolic rate [1, 6]. Moreover, it is rich in 4
lipid with unsaturated fatty acid and has only the moderate amount of antioxidant enzymes [7]. Thus, the brain is more vulnerable to iron- and ROS-induced toxicity.
Because both iron overload and iron deficiency causes neuronal dysfunction, the brain expresses several iron management proteins, which are involved in the uptake, export, storage and utilization of the iron [13], to regulate its iron content. The proteins involved in the brain iron homeostasis system include HFE [11], ferritin, transferrin (Tf), transferrin receptor (TfR) [17], iron regulatory protein (IRP), divalent metal transporter 1
(DMT-1) and cerruloplasmin [6, 9, 12, 18, 19]. The expression of these iron management proteins can be altered in accordance with iron availability in an attempt to maintain the iron content in the brain. For example, iron deficiency results in decreased expression of iron storage protein, ferritin, and increased expression of cellular iron uptake protein, TfR
[20]. The end result would be to reduce iron storage and in the same time to increase iron uptake during iron deficiency. When iron is in excess, ferritin expression is increased while TfR expression is decreased resulting in less iron uptake and increasing iron storage [8, 12].
Traditionally the brain was thought to be protected from iron overload by the blood brain barrier (BBB) which separates and restricts the exchange of iron and iron management proteins between the brain tissue and the blood. Thus, a paradigm became established that the brain was protected from iron-related genetic disorders such as hereditary hemochromatosis (HH). This traditional notion came from misinterpretations of the histochemical studies for iron in the brain in late 1930s and 1940s [21, 22].
However, these studies report that in addition to circumventricular regions, where the
BBB is absent, there was increased deposition of iron observed in other brain regions 5
including the cerebral cortex, hypothalamus, lentiform nucleus and the dentate nucleus that are behind the BBB [21, 22]. Recently, magnetic resonance imaging (MRI) studies have reported iron accumulation in the basal ganglia of patients with HH as well as in the substantia nigra, the red nucleus and the dentate nucleus [23-25]. Therefore, the notion that the brain is protected from iron overload in HH became established in neuroscience despite evidence to the contrary. Iron overload observed in HH is primarily caused by mutations in the HFE gene [10]. The HFE protein resides on the brain microvasculature and choroid plexus where it can impact iron uptake by the brain [26]; therefore, it should not be surprising that HFE mutations are associated with increased brain iron and have recently been proposed to be genetic modifiers for risk of developing neurodegenerative disorders [9, 11, 27]. In addition, mutations in other genes involved in iron metabolic pathway can lead to iron accumulation in brain regions and cause neuronal degeneration.
Neurodegeneration with brain iron accumulation caused by a number of iron-related gene mutations, except HFE gene variants, are summarized in Table 1-1 [9, 12, 13]. This review will cover the current understanding of the association between HFE mutations and neurodegenerative diseases, such as ALS, AD, PD and ischemic stroke.
6
Table 1-1
Neurological diseases with Genes Functions Phenotypes brain iron accumulation
Facilitates iron transport by Aceruloplasminemia caused by Ataxia, cognitive Ceruloplasmin oxidizing ferrous to ferric autosomal recessive impairment, retinal iron cerruloplasmin gene mutation degeneration
Mitochondrial protein with Friedreich’s ataxia caused by ferroxidase activity, which Progressive cerebellar reduction in frataxin protein Frataxin involves in heme protein ataxia, muscle weakness due to GAA triple-repeat synthesis and iron-sulfur and sensory loss expansion in frataxin gene cluster biosynthesis
Neuroferritinopathy caused by Limb dystonia, autosomal dominant mutations spasticity and rigidity, L-ferritin Iron storage protein in gene encoding L-chain and subtle cognitive ferritin impairment
Pantothenate A key enzyme for Hallervorden-Spatz disease Dystonia, dysarthria, kinase 2 mitochondrial coenzyme A caused by autosomal recessive rigidity and pigmentary (PANK2) biosynthesis PANK2 mutation retinopathy
Progressive memory The C2 variant of transferrin and cognitive decline, Transferrin Transports iron to neurons gene is a proposed genetic risk and most common type factor for Alzheimer’s disease of dementia
Table 1-1. Neurodegeneration with brain iron accumulation caused by mutations in genes involved in iron metabolism 7
1.4. The HFE Gene
Simon et al. [28] first demonstrated in late 1970s that the gene responsible for hereditary hemochromatosis (HH) is closely linked to the human leukocyte antigen
(HLA) locus on a short arm of chromosome 6. Twenty years later, this gene was identified and termed as HLA-H gene by Feder et al [29]. The HLA-H gene, now renamed HFE gene, is comprised of seven exons and is expressed widely or at low level in most of the tissues including brain.
Two common polymorphisms in the HFE gene associated with HH. A G-to-A transition at nucleotide 845 changes cysteine to tyrosine at amino acid 282 (C282Y) and a
C-to-G transition at nucleotide 187 results in a histidine to aspartic acid substitution at amino acid 63 (H63D) [29]. Eighty five to ninety percent of HH patients are homozygous for C282Y variant and 5 % of patients are compound heterozygous for C282Y and H63D variants [29-31]. HH is an autosomal recessive disorder characterized by an excessive absorption of dietary iron leading to abnormal iron accumulation with secondary tissue damage in various organs such as liver, pancreas and heart. The clinical consequences of
HH include cirrhosis, hepatomegaly, diabetes mellitus and cardiomyopathy [32, 33]. HH is the most common inherited disorder in individuals of Northern European descent with
1 in every 200-400 individuals with even higher prevalence in Ireland (1:100) [30].
Although the C282Y HFE variant is more commonly associated with HH, the H63D HFE variant is present more frequently in general population with worldwide frequency of
8.1% compared to the C282Y variant (1.9%). Similar to the distribution of HH, C282Y
HFE variant is present more abundantly in Northern European population and Northern
European descent. The H63D HFE variant has a more general and broader distribution 8
with a higher frequency in Europe (14.9%) and moderate frequency in Asia, Africa,
Middle East and America [30, 34]. The product of the HFE gene is a 343 residue type 1 transmembrane glycoprotein named HFE protein [29].
1.5. The Structure of HFE Protein
The HFE protein is a member of the major histocompatibility complex class-1
(MHC-1)-like family. Like other MHC class-I like molecules, a single polypeptide HFE protein contains a transmembrane region, short cytoplasmic tail and 3 extracellular domains (α1, α2 and α3) [29, 35]. Two peptide-binding domains (α1 and α2) consist of 8 antiparallel β strands and 2 antiparallel α helices, and position on the top of an immunoglobulin-like domain (α3) [35]. The HFE protein contains four cysteine residues forming a disulfide bridge in α2 and α3 domains [29], which is one of the important conserved features of MHC class-I family required for non-covalent interacting with β2 microglobulin [36] and for transport from the endoplasmic reticulum (ER) to the cell surface [37, 38]. However, the peptide binding groove in the HFE protein is narrower than other MHC-1 proteins due to the translation of α1 helix toward α2 helix. Moreover, the HFE protein has only 2 of 4 tyrosine residues in the peptide-binding region, which are important for peptide binding. Therefore, unlike other MHC-1 proteins, the HFE protein does not function as an antigen-presenting molecule [29, 39].
1.6. Function of HFE Protein
The major function of the HFE protein is to regulate iron homeostasis. The HFE protein interacts with β2 microglobulin in the ER and is transported to the plasma 9
membrane [38, 40] where the HFE protein forms a stable complex with the TfR [41]. The binding of the HFE protein to the TfR is pH dependent with a tight interaction at pH 7.5 but very weak or no binding at acidic pH [39, 42]. The HFE-TfR interaction lowers the affinity of the TfR for holotransferrin, i.e., Fe-Tf [41]. Lebrón et al. later reported that
HFE bound to TfR at or near the Fe-Tf binding site where it could sterically hinder Tf binding to the TfR [43]. In the absence of the HFE, TfR homodimers bind two Fe-Tf molecules while HFE bound TfR binds only one Fe-Tf molecule by forming a ternary complex consisting of one HFE, two TfR polypeptide chain and one Fe-Tf [39, 43].
Therefore, the HFE protein functions in regulation of iron homeostasis by binding to the
TfR and reducing the transport of Fe-Tf molecules (Figure 1-1).
The cysteine residue in α3 domain that is altered in the C282Y HFE variant is one of the conserved residues important for forming disulfide linkages and interacting with
β2-microgloblulin for cell surface expression [29, 38, 40]. Therefore, the C282Y HFE variant is located primarily intercellular [38, 40] and does not bind to the TfR to limit transferrin-mediated iron uptake [41]. The functional consequence of the H63D HFE variant was first reported by Feder and colleagues [41]. Similar to the wild-type HFE protein, the H63D HFE interacts with β2-microglobulin and is transported to the plasma membrane [38, 40]; however, the interaction of the H63D HFE with the TfR does not limit transferrin-mediated iron uptake [41]. Since the H63D mutation is present in the α1 domain of the peptide-binding region [29, 43] the functional consequence of this variant is reduction of the affinity of the H63D protein for TfR. Thus, both C282Y and H63D
HFE are associated with an increased iron accumulation compared to expression of the wild-type HFE (Figure 1-1). 10
Figure 1-1
Figure 1-1. Schematic diagram for cellular iron metabolism in the presence of wild-type
(WT) HFE, H63D HFE and C282Y HFE. The WT HFE protein normally interacts with the transferrin receptor (TfR) on the cell surface and regulates transferrin-mediated iron uptake. The H63D HFE interacts with TfR but fails to limit the interaction of TfR with iron-bound transferrin (Tf). The C282Y HFE fails to migrate to the membrane; thus, it does not limit the interaction of TfR with iron-bound transferrin. Therefore, H63D and
C282Y HFE are associated with cellular iron accumulation.
11
In addition to its important role in iron homeostasis, the HFE protein affects a range of cellular functions including innate immunity [44, 45]. Lee et al. [46] developed stable human neuroblastoma cell lines (SH-SY5Y) carrying the wild-type, C282Y or
H63D HFE and demonstrated that HFE mutations were associated with iron accumulation and increased oxidative stress. Neuroblastoma cells carrying HFE mutations, in particular the H63D HFE, also increase intracellular calcium levels, have greater glutamate secretion and reduced uptake [47], increase production of monocyte chemoattractant protein-1 (MCP-1) [48], tau phosphorylation [49] and prolonged ER stress [50]. Each of above mechanisms has been considered as an underlying mechanism contributed to the pathogenesis of neurodegenerative disorders [2, 6, 7, 12, 51].
1.7. HFE and Neurodegenerative Disorders
In order for the HFE protein to impact brain iron accumulation, it should be found in the brain. Indeed the HFE protein is expressed in choroid plexus epithelial cells, endothelial cells of the microvasculature and ependymal cells lining the ventricle in the brain along with TfR, where it can influence iron uptake into the brain [26, 52].
Nevertheless, the relationship between HFE mutations and CNS diseases has not received much attention until recently. Because brain iron concentration increases with age [2, 14] and HFE gene mutations are associated with excess iron accumulation [53-55] in different organs, it is logical that individuals who carry a HFE variant are at higher risk for brain iron accumulation and the accompanying neurological sequelae [33]. A recent study of Bartozokis et al. [56] demonstrated that the presence of the H63D HFE gene variant and/or C2 allele of transferrin gene was associated with increased brain ferritin 12
iron in older men compared to non-carriers. Given the presence and location of the HFE protein at the interface between the brain and the vasculature, and the CSF where it can influence brain iron uptake [26, 52] it is not a surprise that the mutant forms of HFE could contribute to iron overload in neurodegenerative disorders [6, 12, 19, 57-60]. Thus, it was a logical quest to examine how HFE genotypes influence the course of neurodegenerative disorders.
1.7.1. HFE and Amyotrophic Lateral Sclerosis
The association between polymorphisms in the HFE gene and amyotrophic lateral sclerosis (ALS) was first reported by Wang et al. [61]. The H63D HFE was present in as many as 30% of ALS patients, which was higher than that reported for the superoxide dismutase (SOD1) mutation [62] and represented the second most frequent genetic variation found in ALS [61]. Subsequently, four other groups in United Kingdom [63],
Italy [64], the Netherlands [65] and China [66] have reported an increased incidence of
H63D HFE in patients with ALS. Moreover, meta-analysis including 66,000 cases revealed that homozygosity for H63D HFE was associated with 4-fold risk of developing
ALS [27]. In contrast to those studies, Yen et al. [67] and two recent studies by Praline et al. [68] and van Rheenen et al. [69] reported that H63D HFE is not associated with ALS, age of disease onset or disease progression. One of the studies [67] was limited by a low number of ALS patients and ~10% of those ALS patients represented ethnic backgrounds in which HFE mutations are absent or present at lower frequency than the general population [34]. Despite the findings of no association between H63D HFE and ALS in 3 out of 8 studies, most of those studies, except the study in Chinese population [66], 13
reported that H63D HFE is present in approximately 25 to 34% of ALS patients. Given the consideration that worldwide frequency for H63D is 8.1% [30], findings of H63D
HFE in as many as 34% of ALS patients strongly indicate that H63D HFE contributes to
ALS. Even in Chinese population in which H63D HFE is present at lower frequency
(2.8%; [66]), 10% of ALS patients carry H63D HFE. This finding further enhances the idea that H63D HFE is the risk factor for ALS. In contrast to H63D HFE, no association has been identified for C282Y HFE with ALS [27, 61, 63-65, 67, 68]. Findings from studies evaluating an association between HFE mutations and ALS are summarized in
Table 1-2. 14
Table 1-2
Number of Location ALS (%) Control (%) Comments Refs subjects
121: ALS Pennsylvania H63D: 29.75 H63D: 14.29 Higher frequency of H63D in [61] 133: Control (USA) C282Y: 0.83 C282Y: 0 sporadic ALS patients
Birmingham 379: ALS H63D: 34.3 H63D: 22 Overrepresentation of H63D in [63] and Ireland 400: Control C282Y: 18.2 C282Y: 19 sporadic ALS (UK)
149: ALS Torino H63D: 28.8 H63D: 14.8 Increased incidence of H63D [64] 168: Control (Italy) C282Y: 3.3 C282Y: 1.8 in sporadic ALS patients
Homozygosity for H63D [65] 289: ALS Utrecht H63D: 27.8 H63D: 26.7 predisposed to sporadic ALS 5886: Control (Netherlands) C282Y: 8.3 C282Y: 11.5 and higher age at onset in H63D carriers
195: ALS Higher risk for ALS in H63D [66] China H63D: 10.3 H63D: 3.2 405: Control carriers
No association between HFE [67] genotypes and the age at onset, 51: ALS Texas H63D: 25.5 H63D: 23.4 rate of progression in sporadic 47: Control (USA) C282Y: 3.9 C282Y: 4.3 ALS (Small samples and ethnically diverse cases and controls)
824: ALS H63D: 27.0 H63D: 30.2 H63D polymorphism is not [68] France 447: Control C282Y: 6.3 C282Y: 11.6 associated with sporadic ALS
H63D is not associated with [69] 3962: ALS Europe H63D: 26.0 H63D: 27.0 ALS, age at disease onset or 5072: Control survival
Table 1-2. Studies evaluating an association between HFE mutations and amyotrophic lateral sclerosis. % represents the genotype frequency for HFE mutations among ALS patients and controls respectively. 15
Iron dysregulation caused by mutations in the HFE gene is likely to contribute to the relationship between the H63D HFE and ALS. A number of studies have reported increased iron levels in the central nervous system (CNS) of ALS patients [57, 58, 70,
71]. In the ALS animal models, iron chelation treatment delayed disease onset, extended survival and prevented neuronal degeneration [72, 73] suggesting iron metabolism contributed to the disease process. Sporadic ALS patients have elevated serum ferritin
[74, 75], which was even more increased in ALS patients with the H63D gene variant
[76]. Cell culture models developed to identify the relationship of HFE gene variants to pathogenic factors in ALS have also identified oxidative stress [46], glutamate release and decreased uptake [47], increased MCP-1 secretion suggesting an alteration in innate immunity [48], increased tau phosphorylation [49] and ER stress [50] in H63D carrying cells compared to the wild-type. It is probable that these mechanisms converge to mediate motor neuron toxicity in ALS and we have proposed that H63D HFE creates a permissive milieu for ALS pathogenic factors.
In addition to being a risk factor for ALS, the HFE gene variant could also impact treatment of ALS. For example, dysphagia is a common symptom in ALS patients leading to increased risk of malnutrition, weight loss and dehydration; therefore, patients will receive an enteral nutrition treatment [77, 78]. The iron content of enteral formulas ranges from 13-24 mg/L and most patients will receive at least 1 L/day, which is higher than recommended daily iron intake (10 mg/day). Enteral nutrition regimens can last from 30 to 1460 days, depending on the rate of disease progression [79]. Since the HFE mutation is associated with iron overload the long-term enteral nutrition treatment in ALS patients with HFE mutations may worsen iron accumulation. Moreover, it has been 16
reported that elevated serum ferritin indicative of increased body iron storage is negatively associated with survival in ALS patients [75]. Therefore, stratification by HFE genotypes should be considered before initiating the long-term enteral nutrition treatment in ALS patients.
Moreover, ALS patients with HFE mutations may respond differently to therapeutic interventions compared to those without HFE mutations. For example, minocycline, an antibiotic with iron chelation, anti-oxidant, anti-inflammatory and mitochondria-protective properties, had no benefit in ALS patients [80] despite its beneficial effect in rodent ALS models [81-84]. Mitchell et al. [48] demonstrated that minocycline reduced MCP-1 release in H63D, but not in wild-type HFE, expressing neuroblastoma cell lines (SH-SY5Y). Therefore, patients’ heterogeneity for HFE alleles may explain the lack of benefit of minocycline in ALS patients. Because the H63D HFE variant is present in approximately 30% of ALS patients [61] HFE genotypes should be considered when assessing treatment strategies or therapeutic interventions for ALS. We have recently found that a mouse line carrying the H67D HFE gene variant (the equivalent of the human H63D variant) when crossed with the SOD1 mouse model for
ALS results in a shorter survival and shorter disease duration (Chapter 3). This exciting observation extends the cell culture observations to an in vivo model.
1.7.2. HFE and Alzheimer’s Disease
Excessive iron accumulation in neuritic plaques and neurofibrillary tangles, and oxidative stress [11, 12, 26, 85-88] are consistent observations in pathogenesis of
Alzheimer’s disease (AD). Pathological features associated with AD such as senile 17
plaque formed by amyloid beta peptide, neurofibrillary tangle (NFT), hyperphosphorylation of tau protein, can be influenced by iron. For example, iron modulates the ability of α-secretase to cleave amyloid precursor protein (APP) [89], promotes A-β toxicity [90] and aggregation, and can directly regulate the synthesis and expression of APP via the iron responsive element at 5' untranslated region of APP mRNA [88, 91, 92].
Given the involvement of the HFE protein in iron regulation and oxidative stress, mutations in the HFE gene can be expected to impact AD pathogenesis. This statement is supported by studies demonstrating that the HFE protein is expressed in glial and neuronal cells associated with neuritic plaques and NFT [11, 26]. In addition, HFE can be induced by stress factors such as serum deprivation and β-amyloid in mouse microglia
BV-2 cells. Induction of HFE decreases labile iron pool, which may be a protective response to limit iron uptake during cellular stress [93]. The mutant forms of HFE protein, however, may not limit iron uptake effectively or may alter iron metabolism, which would result in intracellular iron overload and increased cell vulnerability to oxidative stress.
Moreover, studies by Hall et al. [94, 95] suggest an association between the H63D
HFE with tau phosporylation and Prolyl-peptidyl isomerase (Pin1), an enzyme responsible for phosphorylation of APP and tau. The expression of the H63D HFE up- regulated the tau phosphorylation in neuroblastoma cell lines by decreasing Pin1activity
[94] and increasing glycogen synthase kinase (GSK)-3β activity [95]. The H63D effect appears to be associated with iron and subsequent oxidative stress because iron exposure increased tau phosphorylation while anti-oxidant Trolox, vitamin E analog, treatment 18
decreased tau phosphorylation [95]. Iron chelation with desferroxamine (DFO) and
Trolox exposure decreased Pin1 phosphorylation. Consistent with the cell model study, a
H67D knock-in mouse line (mouse homologous to H63D in human population) also exhibited increased in Pin1 phosphorylation [94]. All of the above findings suggest that the basic biochemistry of cells is altered in the presence of the H63D variant and provide a compelling proposition that HFE gene variants are modifying risk factors for AD.
During the past 10 years, multiple studies have addressed the association of the
HFE gene mutations with AD. Moalem and colleagues [96] were the first to report that the HFE mutations were overrepresented in males with familial AD (FAD) and among non-carriers of ApoEε4 allele, a well-known genetic risk factor for AD. The HFE mutations predisposed males to FAD but were somewhat protective in females. A study conducted by Percy et al. [97] found that in folate-supplemented Ontario population, the presence of both ApoEε4 and H63D HFE predisposed females to AD but not males.
Robson et al. [98] showed bi-carriers of C282Y HFE and C2 allele of transferrin, both involved in iron metabolism, to have 5 times greater risks of AD. Pulliam et al. [99] included individuals with mild cognitive impairment (MCI) and non-demented controls with AD-like pathology (HPC) in addition to AD patients in their study and reported that proportion of homozygous or compound heterozygous for HFE mutations was higher in
AD/MCI and HPC patients. They extended their study by evaluating ventricular (CSF) fluid F2-isoprostane level, which hallmarks brain lipid peroxidation, and found the association of HFE mutations with increased oxidative stress in AD patients [99].
Sampietro et al. [100] examined the effect of HFE mutations on age of onset in patients with sporadic AD and found that patients who were either heterozygous or 19
homozygous for the H63D HFE developed AD on average 5 years earlier than those with wild-type HFE. In that same study, in patients who develop disease symptoms before 70 years old, H63D variant was found twice as frequently than in patients who develop AD between 70 and 80 years old and 5 times more frequently than in patients who develop
AD after 80 years old. Sampietro et al. [100] also suggested that the effect of HFE alleles on age of onset is independent of ApoEε4 genotypes in Italian AD patients, but
Combarros and colleagues [101] found the synergistic effect of the H63D HFE and
ApoEε4 on age of onset in Spanish AD patients. Combarros et al. reported that the presence of one or 2 copies of the H63D allele in ApoEε4 homozygotes significantly reduced the age of onset when compared to ApoEε4 heterozygotes or non-carriers of
ApoEε4 [101]. Recently we demonstrated that presence of H63D HFE, similar to
ApoEε4, decreases brain cholesterol levels in H67D knock-in mice. Altered brain cholesterol is accompanied by greater reduction in brain volume and poorer recognition and spatial memory [102]; phenotypes associated with AD. Given that both H63D HFE and ApoEε4 disrupt brain cholesterol metabolism, excessive loss of brain cholesterol levels in the presence of both mutations could be associated with an early age of AD onset. Similarly, Correia et al. [103] found an association of the H63D HFE with earlier age of onset in Portuguese AD patients. However, their study showed the negative association between C282Y HFE with AD suggesting the protective role of C282Y HFE in AD. Meta-analysis including 66,000 cases and 226,000 controls showed a weak association (1.1 fold risk) for AD in individuals with H63D/H63D [27]. Studies reporting an association between HFE mutations and AD are summarized in Table 1-3.
20
Table 1-3
Number of Refs Location AD (%) Control (%) Comments subjects
26: FAD Toronto H63D: 26.9 H63D 26.8 Higher frequency of HFE mutations [96] 41: Control Ontario C282Y: 15.4 C282Y: 9.8 in males with FAD
Male Male 54: SAD Toronto [97] H63D: 27.3 H63D: 52 58: Control Ontario C282Y: 9.1 C282Y: 4.0
Female Female Presence of both H63D and E4 27: SAD H63D: 37.5 H63D: 18.2 alleles predisposed females to AD 58: Control C282Y: 18.8 C282Y: 15.2 H63D by itself appeared to protect Male males against AD. (Low frequency Kingston H63D: 22.2 of H63D/H63D in the study partly Ontario C282Y: 0 due to small sample sizes)
Female 27: SAD H63D: 27.8 C282Y: 11.1
191: AD Bi-carriers of C282Y and C2 allele [98] H63D: 27.8 H63D: 27.9 69: MCI Oxford of transferrin are 5 times greater C282Y: 15.7 C282Y: 11.9 269: Control risks of AD
HFE variants HFE variants (homozygous (homozygous More homozygous or compound [99] 133: AD or compound or compound heterozygous HFE mutations in Kentucky 5: MCI heterozygous heterozygous AD/MCI (USA) 67: Control for H63D, for H63D, Increased oxidative stress in AD C282Y and C282Y and patients with HFE mutations S65C): 47.1 S65C): 55.2
H63D mutation was more frequent [100] 107: SAD Milan H63D: 20.6 H63D: 25.3 in SAD patients with an earlier 99: Control (Italy) C282Y: 3.7 C282Y: 4 disease onset
Northern Lower age at AD onset in 328: SAD H63D: 48.2 [101] Spain H63D/ApoE ε4
113: AD Porto H63D: 17.2 H63D: 20.3 H63D is associated with earlier onset [103] 82: Control (Portugal) C282Y: 1.3 C282Y: 5.8 Negative association for C282Y
Basque Higher frequency of H63D in AD [110] 211: AD H63D: 18.0 H63D: 29.9 Country patients 167: Control C282Y: 4.5 C282Y: 3.3 (Spain) 21
Table 1-3. Studies showing an association between HFE mutations and Alzheimer’s disease. % represents the genotype frequency for HFE mutations among AD patients and controls respectively.
22
In contrast to above studies, some studies found no association of HFE alleles with AD. Lleo et al. [104] first reported that neither the C2 nor HFE alleles were associated with an increased risk for AD; however, there was a trend toward a higher frequency for the H63D HFE among male patients with AD. A study by Berlin et al.
[105] showed a similar frequency distribution for HFE alleles among AD and control groups. Although HFE alleles had no significant impact on age of onset or diagnosis, age of onset for cognitive symptoms or severity of neuropsychological deficits, individuals homozygous for H63D HFE displayed a trend toward accelerated conversion from MCI to AD [105]. Candore et al. [106] found neither a significant difference in frequencies of
H63D and C282Y HFE nor the effect of HFE alleles on age of onset between sporadic
AD and controls from Northern Italy. Guerreiro et al. [107] included both AD and MCI patients in their study and found that HFE alleles did not contribute to the risk of developing AD or MCI. In addition, they extended their study by conducting a meta- analysis of both H63D and C282Y HFE in the five published studies regarding AD [96,
98, 100, 105, 106]. They found no significant association between HFE variants and the development of AD [107]. Avila-Gomez et al. [108] found neither the allelic frequencies of H63D HFE mutations nor their effect on age of onset was different between familial
AD patients with E280A mutation in presenilin-1 gene (PSEN-1) and non-demented controls. Alizadeh and colleagues [109] reported no association of HFE alleles with AD, though they did suggest that H63D homozygotes tended to have earlier age of onset compared to non-carriers. Studies reporting no association between HFE variants and AD are summarized in Table 1-4.
23
Table 1-4
Number of Control (%) Location AD (%) Comments Refs subjects
Trend towards an [104] 108: AD H63D: 42.6 H63D: 34.5 Spain increased frequencies of 110: Control C282Y: 3.7 C282Y: 3.6 H63D variants in AD
Trend of an accelerated [105] rate of MCI-to-AD
conversion in H63D 213: AD Montreal H63D: 33 H63D: 34 homozygotes 106: MCI (Canada) C282Y: 5 C282Y: 10 No effect on age at onset 63: Control or diagnosis or onset of cognitive symptoms
HFE variants did not 123: AD H63D: 23.6 H63D: 19.5 [106] Northern Italy influence age at onset or 152: Control C282Y: 1.6 C282Y: 0.7 the risk of AD
HFE genotypes did not [107] contribute to age at onset 130: AD/MCI Coimbra H63D: 34.6 H63D: 35.6 or the risk of AD. (Low 115: Control (Portugal) C282Y: 4.6 C282Y: 4.3 “n” or genetic background)
[108] H63D did not contribute 105: AD H63D: 32.4 H63D: 29.6 Colombia to the risk or age at onset 220: Control C282Y: 0 C282Y: 0.45 of familial AD
Male Male Age at onset tended to be [109] H63D: 32.0 H63D: 28.6 earlier in males 268: AD Rotterdam C282Y: 6.9 C282Y: 11.5 homozygous for H63D 2079: Control (Netherlands) Female Female (no effect of HFE variants H63D: 23.9 H63D: 26.5 on the risk of AD) C282Y: 8.7 C282Y: 13.0
Table 1-4. Studies showing no association between HFE mutations and Alzheimer’s disease. % represents the genotype frequency for HFE mutations among AD patients and controls respectively. 24
Blaquez et al. [110] were the first to report the negative association between
H63D HFE and AD suggesting the protective role of this gene variant in AD but they did not observe any difference between AD and controls regarding C282Y HFE. Similarly, a recent meta-analysis of 22 studies including 4,365 cases and 8,652 controls reported that
H63D HFE is protective for risk of AD, particularly in Caucasians though no significant association was found for C282Y HFE with AD [111].
Population differences in the frequencies of HFE alleles and the interaction with environmental factors, age, gender, and/or other genes may all attribute to these conflicting findings regarding to the association of the HFE gene variants with complex diseases like AD. Animal models involving mutations in the HFE gene are required to further elucidate the association between HFE gene variants and AD independent of environmental conditions which are significant confounding variables in the human studies. The effect of the gene-environmental interaction on the neurological consequences of HFE gene variants and the importance of animal models involving HFE mutations will be discussed later in this review.
1.7.3. HFE and Parkinson’s Disease
Excess accumulation of iron in the substantia nigra, detected by postmortem examination of brain tissues from Parkinson’s (PD) patients and MRI [112], is a consistent observation in PD [12, 19, 88, 113]. Free iron promotes Parkin and α-synuclein aggregation; thus enhances the generation of Lewy-body, a pathological hallmark of PD
[12, 19, 113]. Nielsen et al. [23] reported the relationship between iron accumulation in the basal ganglia and development of a parkinsonian syndrome in a HH patient. In 25
addition, a recent case study reported 4 patients with concurrent hereditary hemochromatosis (HH) and idiopathic Parkinson’s disease (IPD), and the authors suggested that increased iron level in the basal ganglia could be associated with symptoms of IPD [114]. These studies all suggest the role of iron in pathophysiology of
PD and the HFE gene mutations, associated with iron accumulation, are genetic risk factors for PD.
The studies that have sought to identify a more direct link between HFE mutations and PD have given mixed results. A study in Portuguese population found an increased frequency of C282Y carriers in PD but its presence did not impact age of onset
[107]. Dekker et al. [115] determined the role of HFE mutations in PD and parkinsonism in two population-based series from the Netherlands. They found an increase in frequency of the C282Y variant in PD compared to controls. Moreover, more patients with parkinsonism were carriers for the C282Y variant in both populations. The frequency of the H63D variant between PD and controls was not different [115]. A study in Australia, however, reported that the presence of the C282Y HFE was protective against the development of PD [116]. Findings from studies reporting an association of the C282Y HFE variant with PD are summarized in Table 1-5.
26
Table 1-5
Number of Location PD (%) Control (%) Comments Refs subjects
132: PD Coimbra H63D: 32.6 H63D: 35.6 Higher prevalence of C282Y [107] 115: Control (Portugal) C282Y: 13.6 C282Y: 4.3 carriers in PD patients
137: PD PD [115] 47: non-PD PS H63D: 26.3 Increased frequencies of
2914: Control C282Y: 10.9 C282Y homozygotes in PD Rotterdam H63D: 26.7 Non-PD PS and increased C282Y C282Y: 11.9 H63D: 19.1 carriers in non-PD PS C282Y: 25.5
60: PD PD Increased frequencies of 25: non-PD PS H63D: 26.7 C282Y carriers in non-PD Southwest 2914: Control C282Y: 10.0 H63D: 26.7 PS Netherlands Non-PD PS C282Y: 11.9 (Small samples for H63D: 20.0 C282Y/C282Y in both C282Y: 32.0 studies)
Protective effect of C282Y [116] 438: PD Queensland C282Y: 10.7 C282Y: 16.5 in PD. (Controls included 485: Control (Australia) patients’ siblings)
Table 1-5. Studies showing an association between the C282Y HFE variant and
Parkinson’s disease. % represents the genotype frequency for HFE mutations among AD patients and controls respectively.
27
Although this review is focused on the H63D and C282Y HFE polymorphisms, a study by Borie et al [117] found those two HFE variants were similar between PD and controls, but that the G258S transferrin polymorphism was present at higher frequency in
PD patients. A group from Germany analyzed the entire coding region of the HFE gene in PD patients [118]. Prior to the analysis of the HFE gene, patients in this study were chosen not only by the clinical symptoms but also by transcranial sonography examination for the subtantia nigra hyperechogenicity, which suggests increased iron level. They identified 2 novel variants of the HFE gene in exon 2 and 4 (K92N and
I217T) present only in PD patients, but no association of the two more common HFE variants, C282Y and H63D variants, with PD [118]. Similarly, the frequencies of the
C282Y and H63D HFE genes were not different between PD and controls in Norwegians
[119] and neither was in the population of Faroe island, where the prevalence of PD and
HFE variants are higher than expected [120]. Biasiotto et al., [121] and more recently
Greco et al. [122] also reported that polymorphisms in HFE gene did not contribute to the risk of PD or clinical features of disease in an Italian population. Studies reporting no association of H63D and C282Y HFE with PD are summarized in Table 1-6.
28
Table 1-6
Number of Location PD (%) Control (%) Comments Refs subjects
No association between HFE [117] genotypes and PD. (Large 216: PD Paris H63D: 36.4 H63D: 33.9 proportion of patients in the 193: Control (France) C282Y: 5.0 C282Y: 5.0 study had positive family history of PD)
Identified rare HFE variants [118] 278: PD H63D: 16.0 H63D: 14.1 (K92N and I217T) in PD Germany 280: Control C282Y: 4.7 C282Y: 5.8 No association of C282Y and H63D with PD
388: PD Central H63D: 18.8 H63D: 21.6 No association between HFE [119] 505: Control Norway C282Y: 14.9 C282Y: 15.2 genotypes and PD
No association between HFE [120] Faroe 79: PD H63D: 31.6 H63D: 28.1 genotypes and PD. Cannot Islands 153: Control C282Y: 15.2 C282Y: 18.3 exclude weak association. (Denmark) (Small sample size)
1 : H63D: 13.13 [121] 475: PD C282Y: 2.0 HFE mutations did not 99: Control 1 Milan H63D: 14.53 2 : H63D: 11.2 influence the development 152: Control 2 (Italy) C282Y: 1.7 C282Y: 0.3 of the clinical features of PD 2100:Control 3 3 : H63D: 13.3 C282Y: 1.6
HFE polymorphisms did not 181: PD Calabria H63D: 31.0 H63D: 21.7 [122] contribute to the risk of PD. 180: Control (Italy) C282Y: 1.7 C282Y: 1.7 (insufficient power of study)
Table 1-6. Studies showing no association of H63D and C282Y HFE with Parkinson’s disease. % represents the genotype frequency for HFE mutations among AD patients and controls respectively. 29
Collectively, these studies reported no association for H63D HFE and PD; whereas, the association between C282Y HFE and PD remains inconclusive. The smaller sample size and inappropriate control or patient samples may contribute to divergent results of previous studies. For example, a study in the French population [122] included a large proportion of patients with positive family history of PD. In patients with family history of PD, genes other than HFE may have a stronger effect on the development of
PD; therefore, including patients with positive family history of PD in a study may underestimate the effect of the HFE mutations on PD. Moreover, conflicting findings from above studies concerning the role of the HFE mutations in PD is most likely a result of an interaction between genes and environmental factors such as diet. It has been reported that high dietary intake of iron was associated with risk for PD [123, 124].
Therefore, animal models with HFE mutations, in particular C282Y HFE, will provide evidence-based argument for a relationship between HFE mutations and PD independent of environmental conditions.
1.7.4. HFE and Ischemic Stroke
Following severe ischemic-anoxic insult and subsequent resuscitation, increased iron accumulation was detected in the basal ganglia, thalami, and periventricular and subcortical white matter by MRI [125]. Also a correlation between iron concentration in the basal ganglia of acute ischemic stroke patients, detected by T2* MRI, and stroke lesion growth has been reported (discussed in [51]). The population-base study in the
Netherlands found a significant association between higher serum ferritin concentrations, an indication of high body iron stores, with increased risk for ischemic stroke in 30
postmenopausal women [126]. Similarly, plasma and CSF ferritin levels were higher in patients with progressive stroke compared to those with stable stroke [127] and high baseline serum ferritin level was associated with the poor clinical outcome in patients with acute ischemic stroke [128, 129]. In vivo studies also suggest that brain ischemia disrupts iron homeostasis system resulting in an excess iron deposition in the brain, which contributes to lipid peroxidation of the cell membrane and neuronal death [51].
The catalytic role of iron in the production of reactive hydroxyl radical via a Fenton reaction may explain the link between iron and ischemic stroke.
Because dysregulation of iron homeostasis is associated with neuronal damage following ischemic stroke and iron plays a role in platelet activation, atherosclerosis, diabetes mellitus and hypercholesterolemnia, which are pre-disposing factors to ischemic stroke [51], the HFE gene mutations associated with iron overload could be expected to influence the risk of ischemic stroke. However, the relationship between the HFE gene and ischemic stroke has received little attention and studies concerning the role of HFE variants in ischemic stroke are inconclusive.
Two prospective studies have reported the association between the HFE gene variants and stroke. The first study reported that homozygosity for H63D HFE increased the risk of ischemic cerebrovascular disease and ischemic stroke by 2 to 3 fold [130]. The second study reported an increased risk of cerebrovascular death in women who were carriers for the C282Y HFE. The mortality rate for cerebrovascular death in C282Y carriers was further increased by smoking and hypertension [131]. This finding suggests a potentially critical connection between HFE gene mutations and other risk factors of ischemic stroke such as smoking and hypertension [131, 132]. This notion is further 31
supported by the findings demonstrating that the risk for stroke in HFE carriers (H63D and C282Y) is increased to 2.6 and 3 fold respectively, in patients with hypertension or in individuals who smoke although a weak association (1.3 fold) was observed for HFE genotypes and stroke in the absence of other risk factors [132].
In contrast to the above studies, a group from Sweden conducted the prospective study in which the role of both iron status and HFE genotypes in ischemic stroke were studied simultaneously and they found that neither the C282Y and H63D HFE nor high iron stores influenced the risk of ischemic stroke [133].
Although studies into the relationship between HFE mutations and ischemic stroke have given mixed results, more studies have reported evidence for H63D and
C282Y HFE to impact the disease, particularly in the presence of other risk factors for stroke, such as smoking. Findings from studies evaluating an association of HFE variants with ischemic stroke are summarized in Table 1-7.
32
Table 1-7
Number of Location Stroke (%) Control (%) Comments Refs subjects
Homozygosity for H63D [130] predicts a 2- to 3-fold risk 393: Stroke H63D: 25.7 H63D: 23.6 Denmark of ischemic cerebrovascular 8577: Control C282Y: 10.2 C282Y: 10.9 diseases and ischemic stroke
79 : Cardiovascular Increased risk of Utrecht [131] cases C282Y: 10.7 C282Y: 7.7 cerebrovascular death in (Netherlands) 153: Control female C282Y carriers
HFE mutations increased [132] 202: Stroke Rotterdam H63D: 31.2 H63D: 26.5 risk of stroke only when 2730: Control (Netherlands) C282Y: 14.1 C282Y: 12.2 other risk factors-smoking or hypertension was present
No association between [133] 231: Stroke H63D: 19.1 H63D: 18.9 Sweden HFE mutations and the risk 550: Control C282Y: 8.2 C282Y: 8.9 of ischemic stroke
Table 1-7. Studies evaluating an association between HFE mutations and ischemic stroke.
% represents the genotype frequency for HFE mutations among ALS patients and controls respectively
33
1.8. Animal Models
Because the HFE protein is involved in iron regulation, environmental factors, such as diet, alcohol, and phlebotomy treatment in the case of HH, can be expected to influence the neurological consequences of HFE mutations [31, 134]. These gene- environmental interactions may contribute to inconclusive findings from previous human population studies concerning the relationship between the HFE and neurodegenerative diseases. In order to evaluate the neurological consequences of HFE mutations independent of environmental risk factors that can influence the outcomes of the human population studies, in vivo models are required for study.
The mouse homolog for the human HFE gene has been identified by Hashimoto et al. [135] and it has ~66% amino acid sequence homology to the human HFE gene. An
HFE knock-out mouse line was generated by Zhou et al. [136] to evaluate whether the deficiency or loss of the function of HFE gene product is an underlying mechanism for
HH. Similar to the biochemical abnormalities in HH patients, the HFE knockout mice exhibited excessive hepatic iron concentrations and higher transferrin saturation than the wild-type mice [136]; however, brain iron accumulation was not detected in HFE knockout mice [137, 138]. Tomatsu and colleagues [139] later generated the HFE knock- in mice: H67D (mouse homolog for H63D in humans) and C294Y (mouse homolog for
C282Y in humans) knock-in mice. Hepatic iron concentration in H67D and C294Y knock-in mice was higher than wild-type but lower than HFE knockout mice. Since the consequences of the HFE mutations may result from gain of function of HFE mutations, in vivo models carrying analogous HFE mutations are likely to be better models for evaluating the neurological consequences of the HFE mutations. 34
Understanding the role of the HFE mutations in neurodegenerative diseases has an important clinical implication. Because HFE mutations can influence iron level, inflammatory responses and oxidative stress, individuals who carry HFE mutations may have a higher baseline stress, iron concentration and inflammation than individuals with wild-type HFE. Thus, it can be expected that HFE genotypes will influence responses to anti-oxidant, metal chelators and anti-inflammatory treatments. Therefore, stratification of patient populations by HFE genotypes should be included when evaluating treatment strategies. The HFE knock-in mice may also prove useful in testing how intervention strategies may be impacted by HFE genotypes.
35
1.9. References
[1] B.H. Wigglesworth JM, Iron dependent enzymes in the brain, in: M.B. Youdim (Ed.)
Brain Iron: neurochemical and behavioural aspects, Taylor and Francis, New York,
1988, pp. 25-66.
[2] J.L. Beard, J.R. Connor, B.C. Jones, Iron in the brain, Nutr Rev, 51 (1993) 157-170.
[3] J.R. Connor, S.L. Menzies, Relationship of iron to oligodendrocytes and myelination,
Glia, 17 (1996) 83-93.
[4] B. Todorich, J.M. Pasquini, C.I. Garcia, P.M. Paez, J.R. Connor, Oligodendrocytes
and myelination: the role of iron, Glia, 57 (2009) 467-478.
[5] J.L. Beard, Why iron deficiency is important in infant development, J Nutr, 138
(2008) 2534-2536.
[6] G.A. Salvador, Iron in neuronal function and dysfunction, Biofactors, 36 (2010) 103-
110.
[7] B. Halliwell, Reactive oxygen species and the central nervous system, J Neurochem,
59 (1992) 1609-1623.
[8] C.W. Levenson, N.M. Tassabehji, Iron and ageing: an introduction to iron regulatory
mechanisms, Ageing Res Rev, 3 (2004) 251-263.
[9] D. Johnstone, E.A. Milward, Molecular genetic approaches to understanding the roles
and regulation of iron in brain health and disease, J Neurochem, (2010).
[10] E. Beutler, V. Felitti, T. Gelbart, N. Ho, Genetics of iron storage and
hemochromatosis, Drug Metab Dispos, 29 (2001) 495-499.
[11] J.R. Connor, S.Y. Lee, HFE mutations and Alzheimer's disease, J Alzheimers Dis,
10 (2006) 267-276. 36
[12] L. Zecca, M.B. Youdim, P. Riederer, J.R. Connor, R.R. Crichton, Iron, brain ageing
and neurodegenerative disorders, Nat Rev Neurosci, 5 (2004) 863-873.
[13] E.E. Benarroch, Brain iron homeostasis and neurodegenerative disease, Neurology,
72 (2009) 1436-1440.
[14] B. Hallgren, P. Sourander, The effect of age on the non-haemin iron in the human
brain, J Neurochem, 3 (1958) 41-51.
[15] S. Aoki, Y. Okada, K. Nishimura, A.J. Barkovich, B.O. Kjos, R.C. Brasch, D.
Norman, Normal deposition of brain iron in childhood and adolescence: MR imaging
at 1.5 T, Radiology, 172 (1989) 381-385.
[16] G. Bartzokis, M. Beckson, D.B. Hance, P. Marx, J.A. Foster, S.R. Marder, MR
evaluation of age-related increase of brain iron in young adult and older normal
males, Magn Reson Imaging, 15 (1997) 29-35.
[17] J.R. Connor, S.L. Menzies, Cellular management of iron in the brain, J Neurol Sci,
134 Suppl (1995) 33-44.
[18] K.J. Thompson, S. Shoham, J.R. Connor, Iron and neurodegenerative disorders,
Brain Res Bull, 55 (2001) 155-164.
[19] S. Altamura, M.U. Muckenthaler, Iron toxicity in diseases of aging: Alzheimer's
disease, Parkinson's disease and atherosclerosis, J Alzheimers Dis, 16 (2009) 879-
895.
[20] Q. Chen, J.R. Connor, J.L. Beard, Brain iron, transferrin and ferritin concentrations
are altered in developing iron-deficient rats, J Nutr, 125 (1995) 1529-1535.
[21] J.H. Sheldon, Hemochromatosis, in, Oxford University Press, London, 1935, pp.
155-159. 37
[22] J. Cammermyer, Deposition of iron in paraventricular areas of the human brain in
hemochromatosis, J Neuropathol Exp Neurol, 6 (1947) 111-127.
[23] J.E. Nielsen, L.N. Jensen, K. Krabbe, Hereditary haemochromatosis: a case of iron
accumulation in the basal ganglia associated with a parkinsonian syndrome, J Neurol
Neurosurg Psychiatry, 59 (1995) 318-321.
[24] H.U. Berg D., Hofmann E., Fischer R., Kraus M., Scheurlen M., Becker G., The
basal ganglia in haemochromatosis, Neuroadiology, 42 (2000) 9-13.
[25] M.P. Rutgers, A. Pielen, M. Gille, Chronic cerebellar ataxia and hereditary
hemochromatosis: causal or coincidental association?, J Neurol, 254 (2007) 1296-
1297.
[26] J.R. Connor, E.A. Milward, S. Moalem, M. Sampietro, P. Boyer, M.E. Percy, C.
Vergani, R.J. Scott, M. Chorney, Is hemochromatosis a risk factor for Alzheimer's
disease?, J Alzheimers Dis, 3 (2001) 471-477.
[27] C. Ellervik, H. Birgens, A. Tybjaerg-Hansen, B.G. Nordestgaard, Hemochromatosis
genotypes and risk of 31 disease endpoints: meta-analyses including 66,000 cases and
226,000 controls, Hepatology, 46 (2007) 1071-1080.
[28] M. Simon, M. Bourel, R. Fauchet, B. Genetet, Association of HLA-A3 and HLA-
B14 antigens with idiopathic haemochromatosis, Gut, 17 (1976) 332-334.
[29] J.N. Feder, A. Gnirke, W. Thomas, Z. Tsuchihashi, D.A. Ruddy, A. Basava, F.
Dormishian, R. Domingo, Jr., M.C. Ellis, A. Fullan, L.M. Hinton, N.L. Jones, B.E.
Kimmel, G.S. Kronmal, P. Lauer, V.K. Lee, D.B. Loeb, F.A. Mapa, E. McClelland,
N.C. Meyer, G.A. Mintier, N. Moeller, T. Moore, E. Morikang, C.E. Prass, L.
Quintana, S.M. Starnes, R.C. Schatzman, K.J. Brunke, D.T. Drayna, N.J. Risch, B.R. 38
Bacon, R.K. Wolff, A novel MHC class I-like gene is mutated in patients with
hereditary haemochromatosis, Nat Genet, 13 (1996) 399-408.
[30] A.T. Merryweather-Clarke, J.J. Pointon, J.D. Shearman, K.J. Robson, Global
prevalence of putative haemochromatosis mutations, J Med Genet, 34 (1997) 275-
278.
[31] G. Weiss, Genetic mechanisms and modifying factors in hereditary
hemochromatosis, Nat Rev Gastroenterol Hepatol, 7 (2010) 50-58.
[32] J. Alexander, K.V. Kowdley, HFE-associated hereditary hemochromatosis, Genet
Med, 11 (2009) 307-313.
[33] K.J. Allen, L.C. Gurrin, C.C. Constantine, N.J. Osborne, M.B. Delatycki, A.J.
Nicoll, C.E. McLaren, M. Bahlo, A.E. Nisselle, C.D. Vulpe, G.J. Anderson, M.C.
Southey, G.G. Giles, D.R. English, J.L. Hopper, J.K. Olynyk, L.W. Powell, D.M.
Gertig, Iron-overload-related disease in HFE hereditary hemochromatosis, N Engl J
Med, 358 (2008) 221-230.
[34] A.T. Merryweather-Clarke, J.J. Pointon, A.M. Jouanolle, J. Rochette, K.J. Robson,
Geography of HFE C282Y and H63D mutations, Genet Test, 4 (2000) 183-198.
[35] M.J. Bennett, J.A. Lebron, P.J. Bjorkman, Crystal structure of the hereditary
haemochromatosis protein HFE complexed with transferrin receptor, Nature, 403
(2000) 46-53.
[36] P.J. Bjorkman, P. Parham, Structure, function, and diversity of class I major
histocompatibility complex molecules, Annu Rev Biochem, 59 (1990) 253-288.
[37] J. Miyazaki, E. Appella, K. Ozato, Intracellular transport blockade caused by
disruption of the disulfide bridge in the third external domain of major 39
histocompatibility complex class I antigen, Proc Natl Acad Sci U S A, 83 (1986) 757-
761.
[38] J.N. Feder, Z. Tsuchihashi, A. Irrinki, V.K. Lee, F.A. Mapa, E. Morikang, C.E.
Prass, S.M. Starnes, R.K. Wolff, S. Parkkila, W.S. Sly, R.C. Schatzman, The
hemochromatosis founder mutation in HLA-H disrupts beta2-microglobulin
interaction and cell surface expression, J Biol Chem, 272 (1997) 14025-14028.
[39] J.A. Lebron, M.J. Bennett, D.E. Vaughn, A.J. Chirino, P.M. Snow, G.A. Mintier,
J.N. Feder, P.J. Bjorkman, Crystal structure of the hemochromatosis protein HFE and
characterization of its interaction with transferrin receptor, Cell, 93 (1998) 111-123.
[40] A. Waheed, S. Parkkila, X.Y. Zhou, S. Tomatsu, Z. Tsuchihashi, J.N. Feder, R.C.
Schatzman, R.S. Britton, B.R. Bacon, W.S. Sly, Hereditary hemochromatosis: effects
of C282Y and H63D mutations on association with beta2-microglobulin, intracellular
processing, and cell surface expression of the HFE protein in COS-7 cells, Proc Natl
Acad Sci U S A, 94 (1997) 12384-12389.
[41] J.N. Feder, D.M. Penny, A. Irrinki, V.K. Lee, J.A. Lebron, N. Watson, Z.
Tsuchihashi, E. Sigal, P.J. Bjorkman, R.C. Schatzman, The hemochromatosis gene
product complexes with the transferrin receptor and lowers its affinity for ligand
binding, Proc Natl Acad Sci U S A, 95 (1998) 1472-1477.
[42] J.A. Lebron, P.J. Bjorkman, The transferrin receptor binding site on HFE, the class I
MHC-related protein mutated in hereditary hemochromatosis, J Mol Biol, 289 (1999)
1109-1118. 40
[43] J.A. Lebron, A.P. West, Jr., P.J. Bjorkman, The hemochromatosis protein HFE
competes with transferrin for binding to the transferrin receptor, J Mol Biol, 294
(1999) 239-245.
[44] C.N. Roy, A.O. Custodio, J. de Graaf, S. Schneider, I. Akpan, L.K. Montross, M.
Sanchez, A. Gaudino, M.W. Hentze, N.C. Andrews, M.U. Muckenthaler, An Hfe-
dependent pathway mediates hyposideremia in response to lipopolysaccharide-
induced inflammation in mice, Nat Genet, 36 (2004) 481-485.
[45] L. Wang, E.E. Johnson, H.N. Shi, W.A. Walker, M. Wessling-Resnick, B.J.
Cherayil, Attenuated inflammatory responses in hemochromatosis reveal a role for
iron in the regulation of macrophage cytokine translation, J Immunol, 181 (2008)
2723-2731.
[46] S.Y. Lee, S.M. Patton, R.J. Henderson, J.R. Connor, Consequences of expressing
mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking
endogenous HFE, Faseb J, 21 (2007) 564-576.
[47] R.M. Mitchell, S.Y. Lee, Z. Simmons, J.R. Connor, HFE polymorphisms affect
cellular glutamate regulation, Neurobiol Aging, (2009).
[48] R.M. Mitchell, S.Y. Lee, W.T. Randazzo, Z. Simmons, J.R. Connor, Influence of
HFE variants and cellular iron on monocyte chemoattractant protein-1, J
Neuroinflammation, 6 (2009) 6.
[49] E.C. Hall, 2nd, S.Y. Lee, N. Mairuae, Z. Simmons, J.R. Connor, Expression of the
HFE allelic variant H63D in SH-SY5Y cells affects tau phosphorylation at serine
residues, Neurobiol Aging, 32 (2011) 1409-1419. 41
[50] Y. Liu, S.Y. Lee, E. Neely, W. Nandar, M. Moyo, Z. Simmons, J.R. Connor, Mutant
HFE H63D protein is associated with prolonged endoplasmic reticulum stress and
increased neuronal vulnerability, J Biol Chem, 286 (2011) 13161-13170.
[51] M.H. Selim, R.R. Ratan, The role of iron neurotoxicity in ischemic stroke, Ageing
Res Rev, 3 (2004) 345-353.
[52] J.M. Bastin, M. Jones, C.A. O'Callaghan, L. Schimanski, D.Y. Mason, A.R.
Townsend, Kupffer cell staining by an HFE-specific monoclonal antibody:
implications for hereditary haemochromatosis, Br J Haematol, 103 (1998) 931-941.
[53] Z.J. Bulaj, L.M. Griffen, L.B. Jorde, C.Q. Edwards, J.P. Kushner, Clinical and
biochemical abnormalities in people heterozygous for hemochromatosis, N Engl J
Med, 335 (1996) 1799-1805.
[54] M.J. Burt, P.M. George, J.D. Upton, J.A. Collett, C.M. Frampton, T.M. Chapman,
T.A. Walmsley, B.A. Chapman, The significance of haemochromatosis gene
mutations in the general population: implications for screening, Gut, 43 (1998) 830-
836.
[55] C. Datz, T. Haas, H. Rinner, F. Sandhofer, W. Patsch, B. Paulweber, Heterozygosity
for the C282Y mutation in the hemochromatosis gene is associated with increased
serum iron, transferrin saturation, and hemoglobin in young women: a protective role
against iron deficiency?, Clin Chem, 44 (1998) 2429-2432.
[56] G. Bartzokis, P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P.
Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely,
J.R. Connor, Prevalent iron metabolism gene variants associated with increased brain
ferritin iron in healthy older men, J Alzheimers Dis, 20 (2010) 333-341. 42
[57] H. Oba, T. Araki, K. Ohtomo, S. Monzawa, G. Uchiyama, K. Koizumi, Y. Nogata,
K. Kachi, Z. Shiozawa, M. Kobayashi, Amyotrophic lateral sclerosis: T2 shortening
in motor cortex at MR imaging, Radiology, 189 (1993) 843-846.
[58] E.J. Kasarskis, L. Tandon, M.A. Lovell, W.D. Ehmann, Aluminum, calcium, and
iron in the spinal cord of patients with sporadic amyotrophic lateral sclerosis using
laser microprobe mass spectroscopy: a preliminary study, J Neurol Sci, 130 (1995)
203-208.
[59] D.A. Loeffler, J.R. Connor, P.L. Juneau, B.S. Snyder, L. Kanaley, A.J. DeMaggio,
H. Nguyen, C.M. Brickman, P.A. LeWitt, Transferrin and iron in normal, Alzheimer's
disease, and Parkinson's disease brain regions, J Neurochem, 65 (1995) 710-724.
[60] S.C. Barber, R.J. Mead, P.J. Shaw, Oxidative stress in ALS: a mechanism of
neurodegeneration and a therapeutic target, Biochim Biophys Acta, 1762 (2006)
1051-1067.
[61] X.S. Wang, S. Lee, Z. Simmons, P. Boyer, K. Scott, W. Liu, J. Connor, Increased
incidence of the Hfe mutation in amyotrophic lateral sclerosis and related cellular
consequences, J Neurol Sci, 227 (2004) 27-33.
[62] M.E. Cudkowicz, D. McKenna-Yasek, P.E. Sapp, W. Chin, B. Geller, D.L. Hayden,
D.A. Schoenfeld, B.A. Hosler, H.R. Horvitz, R.H. Brown, Epidemiology of mutations
in superoxide dismutase in amyotrophic lateral sclerosis, Ann Neurol, 41 (1997) 210-
221.
[63] E.F. Goodall, M.J. Greenway, I. van Marion, C.B. Carroll, O. Hardiman, K.E.
Morrison, Association of the H63D polymorphism in the hemochromatosis gene with
sporadic ALS, Neurology, 65 (2005) 934-937. 43
[64] G. Restagno, F. Lombardo, P. Ghiglione, A. Calvo, E. Cocco, L. Sbaiz, R. Mutani,
A. Chio, HFE H63D polymorphism is increased in patients with amyotrophic lateral
sclerosis of Italian origin, J Neurol Neurosurg Psychiatry, 78 (2007) 327.
[65] N.A. Sutedja, R.J. Sinke, P.W. Van Vught, M.W. Van der Linden, J.H. Wokke,
C.M. Van Duijn, O.T. Njajou, Y.T. Van der Schouw, J.H. Veldink, L.H. Van den
Berg, The association between H63D mutations in HFE and amyotrophic lateral
sclerosis in a Dutch population, Arch Neurol, 64 (2007) 63-67.
[66] X. He, X. Lu, J. Hu, J. Xi, D. Zhou, H. Shang, L. Liu, H. Zhou, B. Yan, L. Yu, F.
Hu, Z. Liu, L. He, X. Yao, Y. Xu, H63D polymorphism in the hemochromatosis gene
is associated with sporadic amyotrophic lateral sclerosis in China, European journal
of neurology : the official journal of the European Federation of Neurological
Societies, 18 (2011) 359-361.
[67] A.A. Yen, E.P. Simpson, J.S. Henkel, D.R. Beers, S.H. Appel, HFE mutations are
not strongly associated with sporadic ALS, Neurology, 62 (2004) 1611-1612.
[68] J. Praline, H. Blasco, P. Vourc'h, V. Rat, C. Gendrot, W. Camu, C.R. Andres, Study
of the HFE gene common polymorphisms in French patients with sporadic
amyotrophic lateral sclerosis, J Neurol Sci, 317 (2012) 58-61.
[69] W. van Rheenen, F.P. Diekstra, P.T. van Doormaal, M. Seelen, K. Kenna, R.
McLaughlin, A. Shatunov, D. Czell, M.A. van Es, P.W. van Vught, P. van Damme,
B.N. Smith, S. Waibel, H.J. Schelhaas, A.J. van der Kooi, M. de Visser, M. Weber,
W. Robberecht, O. Hardiman, P.J. Shaw, C.E. Shaw, K.E. Morrison, A. Al-Chalabi,
P.M. Andersen, A.C. Ludolph, J.H. Veldink, L.H. van den Berg, H63D 44
polymorphism in HFE is not associated with amyotrophic lateral sclerosis, Neurobiol
Aging, 34 (2013) 1517 e1515-1517.
[70] M. Yasui, K. Ota, R.M. Garruto, Concentrations of zinc and iron in the brains of
Guamanian patients with amyotrophic lateral sclerosis and parkinsonism-dementia,
Neurotoxicology, 14 (1993) 445-450.
[71] W.R. Markesbery, W.D. Ehmann, J.M. Candy, P.G. Ince, P.J. Shaw, L. Tandon,
M.A. Deibel, Neutron activation analysis of trace elements in motor neuron disease
spinal cord, Neurodegeneration, 4 (1995) 383-390.
[72] S.Y. Jeong, K.I. Rathore, K. Schulz, P. Ponka, P. Arosio, S. David, Dysregulation of
iron homeostasis in the CNS contributes to disease progression in a mouse model of
amyotrophic lateral sclerosis, J Neurosci, 29 (2009) 610-619.
[73] Q. Wang, X. Zhang, S. Chen, S. Zhang, M. Youdium, W. Le, Prevention of motor
neuron degeneration by novel iron chelators in SOD1(G93A) transgenic mice of
amyotrophic lateral sclerosis, Neuro-degenerative diseases, 8 (2011) 310-321.
[74] E.F. Goodall, M.S. Haque, K.E. Morrison, Increased serum ferritin levels in
amyotrophic lateral sclerosis (ALS) patients, J Neurol, 255 (2008) 1652-1656.
[75] Y. Nadjar, P. Gordon, P. Corcia, G. Bensimon, L. Pieroni, V. Meininger, F.
Salachas, Elevated serum ferritin is associated with reduced survival in amyotrophic
lateral sclerosis, PloS one, 7 (2012) e45034.
[76] R.M. Mitchell, Z. Simmons, J.L. Beard, H.E. Stephens, J.R. Connor, Plasma
biomarkers associated with ALS and their relationship to iron homeostasis, Muscle
Nerve, 42 (2010) 95-103. 45
[77] J.C. Desport, P.M. Preux, C.T. Truong, L. Courat, J.M. Vallat, P. Couratier,
Nutritional assessment and survival in ALS patients, Amyotroph Lateral Scler Other
Motor Neuron Disord, 1 (2000) 91-96.
[78] L.C. Wijesekera, P.N. Leigh, Amyotrophic lateral sclerosis, Orphanet J Rare Dis, 4
(2009) 3.
[79] A. Molfino, I. Kushta, V. Tommasi, F. Rossi Fanelli, M. Muscaritoli, Amyotrophic
lateral sclerosis, enteral nutrition and the risk of iron overload, J Neurol, 256 (2009)
1015-1016.
[80] P.H. Gordon, D.H. Moore, R.G. Miller, J.M. Florence, J.L. Verheijde, C. Doorish,
J.F. Hilton, G.M. Spitalny, R.B. MacArthur, H. Mitsumoto, H.E. Neville, K. Boylan,
T. Mozaffar, J.M. Belsh, J. Ravits, R.S. Bedlack, M.C. Graves, L.F. McCluskey, R.J.
Barohn, R. Tandan, Efficacy of minocycline in patients with amyotrophic lateral
sclerosis: a phase III randomised trial, Lancet Neurol, 6 (2007) 1045-1053.
[81] L. Van Den Bosch, P. Tilkin, G. Lemmens, W. Robberecht, Minocycline delays
disease onset and mortality in a transgenic model of ALS, Neuroreport, 13 (2002)
1067-1070.
[82] J. Kriz, M.D. Nguyen, J.P. Julien, Minocycline slows disease progression in a mouse
model of amyotrophic lateral sclerosis, Neurobiol Dis, 10 (2002) 268-278.
[83] S. Zhu, I.G. Stavrovskaya, M. Drozda, B.Y. Kim, V. Ona, M. Li, S. Sarang, A.S.
Liu, D.M. Hartley, D.C. Wu, S. Gullans, R.J. Ferrante, S. Przedborski, B.S. Kristal,
R.M. Friedlander, Minocycline inhibits cytochrome c release and delays progression
of amyotrophic lateral sclerosis in mice, Nature, 417 (2002) 74-78. 46
[84] W. Zhang, M. Narayanan, R.M. Friedlander, Additive neuroprotective effects of
minocycline with creatine in a mouse model of ALS, Ann Neurol, 53 (2003) 267-270.
[85] M.A. Smith, L.M. Sayre, M.P. Vitek, V.M. Monnier, G. Perry, Early AGEing and
Alzheimer's, Nature, 374 (1995) 316.
[86] M.P. Mattson, J.W. Carney, D.A. Butterfield, A tombstone in Alzheimer's?, Nature,
373 (1995) 481.
[87] M.A. Smith, X. Zhu, M. Tabaton, G. Liu, D.W. McKeel, Jr., M.L. Cohen, X. Wang,
S.L. Siedlak, B.E. Dwyer, T. Hayashi, M. Nakamura, A. Nunomura, G. Perry,
Increased iron and free radical generation in preclinical Alzheimer disease and mild
cognitive impairment, J Alzheimers Dis, 19 (2010) 363-372.
[88] S. Rivera-Mancia, I. Perez-Neri, C. Rios, L. Tristan-Lopez, L. Rivera-Espinosa, S.
Montes, The transition metals copper and iron in neurodegenerative diseases, Chem
Biol Interact, 186 (2010) 184-199.
[89] S. Bodovitz, M.T. Falduto, D.E. Frail, W.L. Klein, Iron levels modulate alpha-
secretase cleavage of amyloid precursor protein, J Neurochem, 64 (1995) 307-315.
[90] C.A. Rottkamp, A.K. Raina, X. Zhu, E. Gaier, A.I. Bush, C.S. Atwood, M. Chevion,
G. Perry, M.A. Smith, Redox-active iron mediates amyloid-beta toxicity, Free Radic
Biol Med, 30 (2001) 447-450.
[91] J.T. Rogers, J.D. Randall, C.M. Cahill, P.S. Eder, X. Huang, H. Gunshin, L. Leiter,
J. McPhee, S.S. Sarang, T. Utsuki, N.H. Greig, D.K. Lahiri, R.E. Tanzi, A.I. Bush, T.
Giordano, S.R. Gullans, An iron-responsive element type II in the 5'-untranslated
region of the Alzheimer's amyloid precursor protein transcript, J Biol Chem, 277
(2002) 45518-45528. 47
[92] H.H. Cho, C.M. Cahill, C.R. Vanderburg, C.R. Scherzer, B. Wang, X. Huang, J.T.
Rogers, Selective translational control of the Alzheimer amyloid precursor protein
transcript by iron regulatory protein-1, J Biol Chem, (2010).
[93] S.Y. Lee, J.R. Connor, Regulation of Hfe by stress factors in BV-2 cells, Neurobiol
Aging, 26 (2005) 803-812.
[94] E.C. Hall, 2nd, S.Y. Lee, Z. Simmons, E.B. Neely, W. Nandar, J.R. Connor, Prolyl-
peptidyl isomerase, Pin1, phosphorylation is compromised in association with the
expression of the HFE polymorphic allele, H63D, Biochim Biophys Acta, 1802
(2010) 389-395.
[95] E.C. Hall, 2nd, S.Y. Lee, N. Mairuae, Z. Simmons, J.R. Connor, Expression of the
HFE allelic variant H63D in SH-SY5Y cells affects tau phosphorylation at serine
residues, Neurobiol Aging, (2009).
[96] S. Moalem, M.E. Percy, D.F. Andrews, T.P. Kruck, S. Wong, A.J. Dalton, P. Mehta,
B. Fedor, A.C. Warren, Are hereditary hemochromatosis mutations involved in
Alzheimer disease?, Am J Med Genet, 93 (2000) 58-66.
[97] M. Percy, S. Moalem, A. Garcia, M.J. Somerville, M. Hicks, D. Andrews, A. Azad,
P. Schwarz, R. Beheshti Zavareh, R. Birkan, C. Choo, V. Chow, S. Dhaliwal, V.
Duda, A.L. Kupferschmidt, K. Lam, D. Lightman, K. Machalek, W. Mar, F. Nguyen,
P.J. Rytwinski, E. Svara, M. Tran, K. Wheeler, L. Yeung, K. Zanibbi, R. Zener, M.
Ziraldo, M. Freedman, Involvement of ApoE E4 and H63D in sporadic Alzheimer's
disease in a folate-supplemented Ontario population, J Alzheimers Dis, 14 (2008) 69-
84. 48
[98] K.J. Robson, D.J. Lehmann, V.L. Wimhurst, K.J. Livesey, M. Combrinck, A.T.
Merryweather-Clarke, D.R. Warden, A.D. Smith, Synergy between the C2 allele of
transferrin and the C282Y allele of the haemochromatosis gene (HFE) as risk factors
for developing Alzheimer's disease, J Med Genet, 41 (2004) 261-265.
[99] J.F. Pulliam, C.D. Jennings, R.J. Kryscio, D.G. Davis, D. Wilson, T.J. Montine, F.A.
Schmitt, W.R. Markesbery, Association of HFE mutations with neurodegeneration
and oxidative stress in Alzheimer's disease and correlation with APOE, Am J Med
Genet B Neuropsychiatr Genet, 119B (2003) 48-53.
[100] M. Sampietro, L. Caputo, A. Casatta, M. Meregalli, A. Pellagatti, J. Tagliabue, G.
Annoni, C. Vergani, The hemochromatosis gene affects the age of onset of sporadic
Alzheimer's disease, Neurobiol Aging, 22 (2001) 563-568.
[101] O. Combarros, M. Garcia-Roman, A. Fontalba, J.L. Fernandez-Luna, J. Llorca, J.
Infante, J. Berciano, Interaction of the H63D mutation in the hemochromatosis gene
with the apolipoprotein E epsilon 4 allele modulates age at onset of Alzheimer's
disease, Dement Geriatr Cogn Disord, 15 (2003) 151-154.
[102] F. Ali-Rahmani, P. Grigson, S. Lee, E. Neely, B. Kinsman, J.R. Connor, C.-L.
Schengrund, Effect of H63D-HFE on cholesterol metabolism, brain atrophy, and
cognitive impairment: implications for Alzheimer Disease, Neurobiology of Aging
(under review).
[103] A.P. Correia, J.P. Pinto, V. Dias, C. Mascarenhas, S. Almeida, G. Porto, CAT53
and HFE alleles in Alzheimer's disease: a putative protective role of the C282Y HFE
mutation, Neurosci Lett, 457 (2009) 129-132. 49
[104] A. Lleo, R. Blesa, C. Angelopoulos, P. Pastor-Rubio, M. Villa, R. Oliva, E. Bufill,
Transferrin C2 allele, haemochromatosis gene mutations, and risk for Alzheimer's
disease, J Neurol Neurosurg Psychiatry, 72 (2002) 820-821.
[105] D. Berlin, G. Chong, H. Chertkow, H. Bergman, N.A. Phillips, H.M. Schipper,
Evaluation of HFE (hemochromatosis) mutations as genetic modifiers in sporadic AD
and MCI, Neurobiol Aging, 25 (2004) 465-474.
[106] G. Candore, F. Licastro, M. Chiappelli, C. Franceschi, D. Lio, C. Rita Balistreri, G.
Piazza, G. Colonna-Romano, L.M. Grimaldi, C. Caruso, Association between the
HFE mutations and unsuccessful ageing: a study in Alzheimer's disease patients from
Northern Italy, Mech Ageing Dev, 124 (2003) 525-528.
[107] R.J. Guerreiro, J.M. Bras, I. Santana, C. Januario, B. Santiago, A.S. Morgadinho,
M.H. Ribeiro, J. Hardy, A. Singleton, C. Oliveira, Association of HFE common
mutations with Parkinson's disease, Alzheimer's disease and mild cognitive
impairment in a Portuguese cohort, BMC Neurol, 6 (2006) 24.
[108] I.C. Avila-Gomez, M. Jimenez-Del-Rio, F. Lopera-Restrepo, C. Velez-Pardo,
Association between HFE 187 C>G (H63D) mutation and early-onset familial
Alzheimer's disease PSEN-1 839A>C (E280A) mutation, Ann Hematol, 87 (2008)
671-673.
[109] B.Z. Alizadeh, O.T. Njajou, M.R. Millan, A. Hofman, M.M. Breteler, C.M. van
Duijn, HFE variants, APOE and Alzheimer's disease: findings from the population-
based Rotterdam study, Neurobiol Aging, 30 (2009) 330-332.
[110] L. Blazquez, D. De Juan, J. Ruiz-Martinez, J.I. Emparanza, A. Saenz, D. Otaegui,
A. Sistiaga, P. Martinez-Lage, I. Lamet, L. Samaranch, C. Buiza, I. Etxeberria, E. 50
Arriola, E. Cuadrado, E. Urdaneta, J. Yanguas, A. Lopez de Munain, Genes related to
iron metabolism and susceptibility to Alzheimer's disease in Basque population,
Neurobiol Aging, 28 (2007) 1941-1943.
[111] M. Lin, L. Zhao, J. Fan, X.G. Lian, J.X. Ye, L. Wu, H. Lin, Association between
HFE polymorphisms and susceptibility to Alzheimer's disease: a meta-analysis of 22
studies including 4,365 cases and 8,652 controls, Molecular biology reports, 39
(2012) 3089-3095.
[112] J.M. Gorell, R.J. Ordidge, G.G. Brown, J.C. Deniau, N.M. Buderer, J.A. Helpern,
Increased iron-related MRI contrast in the substantia nigra in Parkinson's disease,
Neurology, 45 (1995) 1138-1143.
[113] A.M. Snyder, J.R. Connor, Iron, the substantia nigra and related neurological
disorders, Biochim Biophys Acta, 1790 (2009) 606-614.
[114] D.J. Costello, S.L. Walsh, H.J. Harrington, C.H. Walsh, Concurrent hereditary
haemochromatosis and idiopathic Parkinson's disease: a case report series, J Neurol
Neurosurg Psychiatry, 75 (2004) 631-633.
[115] M.C. Dekker, P.C. Giesbergen, O.T. Njajou, J.C. van Swieten, A. Hofman, M.M.
Breteler, C.M. van Duijn, Mutations in the hemochromatosis gene (HFE), Parkinson's
disease and parkinsonism, Neurosci Lett, 348 (2003) 117-119.
[116] D.D. Buchanan, P.A. Silburn, J.B. Chalk, D.G. Le Couteur, G.D. Mellick, The
Cys282Tyr polymorphism in the HFE gene in Australian Parkinson's disease patients,
Neurosci Lett, 327 (2002) 91-94. 51
[117] C. Borie, F. Gasparini, P. Verpillat, A.M. Bonnet, Y. Agid, G. Hetet, A. Brice, A.
Durr, B. Grandchamp, Association study between iron-related genes polymorphisms
and Parkinson's disease, J Neurol, 249 (2002) 801-804.
[118] N. Akbas, H. Hochstrasser, J. Deplazes, J. Tomiuk, P. Bauer, U. Walter, S. Behnke,
O. Riess, D. Berg, Screening for mutations of the HFE gene in Parkinson's disease
patients with hyperechogenicity of the substantia nigra, Neurosci Lett, 407 (2006) 16-
19.
[119] A.H. Aamodt, L.J. Stovner, K. Thorstensen, S. Lydersen, L.R. White, J.O. Aasly,
Prevalence of haemochromatosis gene mutations in Parkinson's disease, J Neurol
Neurosurg Psychiatry, 78 (2007) 315-317.
[120] J. Halling, M.S. Petersen, P. Grandjean, P. Weihe, K. Brosen, Genetic
predisposition to Parkinson's disease: CYP2D6 and HFE in the Faroe Islands,
Pharmacogenet Genomics, 18 (2008) 209-212.
[121] G. Biasiotto, S. Goldwurm, D. Finazzi, S. Tunesi, A. Zecchinelli, F. Sironi, G.
Pezzoli, P. Arosio, HFE gene mutations in a population of Italian Parkinson's disease
patients, Parkinsonism Relat Disord, 14 (2008) 426-430.
[122] V. Greco, E.V. De Marco, F.E. Rocca, F. Annesi, D. Civitelli, G. Provenzano, P.
Tarantino, V. Scornaienchi, F. Pucci, M. Salsone, F. Novellino, M. Morelli, S.
Paglionico, A. Gambardella, A. Quattrone, G. Annesi, Association study between
four polymorphisms in the HFE, TF and TFR genes and Parkinson's disease in
southern Italy, Neurological sciences : official journal of the Italian Neurological
Society and of the Italian Society of Clinical Neurophysiology, 32 (2011) 525-527. 52
[123] C.C. Johnson, J.M. Gorell, B.A. Rybicki, K. Sanders, E.L. Peterson, Adult nutrient
intake as a risk factor for Parkinson's disease, Int J Epidemiol, 28 (1999) 1102-1109.
[124] K.M. Powers, T. Smith-Weller, G.M. Franklin, W.T. Longstreth, Jr., P.D.
Swanson, H. Checkoway, Parkinson's disease risks associated with dietary iron,
manganese, and other nutrient intakes, Neurology, 60 (2003) 1761-1766.
[125] R.B. Dietrich, W.G. Bradley, Jr., Iron accumulation in the basal ganglia following
severe ischemic-anoxic insults in children, Radiology, 168 (1988) 203-206.
[126] A.D. van der, D.E. Grobbee, M. Roest, J.J. Marx, H.A. Voorbij, Y.T. van der
Schouw, Serum ferritin is a risk factor for stroke in postmenopausal women, Stroke,
36 (2005) 1637-1641.
[127] A. Davalos, J. Castillo, J. Marrugat, J.M. Fernandez-Real, A. Armengou, P.
Cacabelos, R. Rama, Body iron stores and early neurologic deterioration in acute
cerebral infarction, Neurology, 54 (2000) 1568-1574.
[128] A. Davalos, J.M. Fernandez-Real, W. Ricart, S. Soler, A. Molins, E. Planas, D.
Genis, Iron-related damage in acute ischemic stroke, Stroke, 25 (1994) 1543-1546.
[129] M. Millan, T. Sobrino, M. Castellanos, F. Nombela, J.F. Arenillas, E. Riva, I.
Cristobo, M.M. Garcia, J. Vivancos, J. Serena, M.A. Moro, J. Castillo, A. Davalos,
Increased body iron stores are associated with poor outcome after thrombolytic
treatment in acute stroke, Stroke, 38 (2007) 90-95.
[130] C. Ellervik, A. Tybjaerg-Hansen, M. Appleyard, H. Sillesen, G. Boysen, B.G.
Nordestgaard, Hereditary hemochromatosis genotypes and risk of ischemic stroke,
Neurology, 68 (2007) 1025-1031. 53
[131] M. Roest, Y.T. van der Schouw, B. de Valk, J.J. Marx, M.J. Tempelman, P.G. de
Groot, J.J. Sixma, J.D. Banga, Heterozygosity for a hereditary hemochromatosis gene
is associated with cardiovascular death in women, Circulation, 100 (1999) 1268-
1273.
[132] O.T. Njajou, M. Hollander, P.J. Koudstaal, A. Hofman, J.C. Witteman, M.M.
Breteler, C.M. van Duijn, Mutations in the hemochromatosis gene (HFE) and stroke,
Stroke, 33 (2002) 2363-2366.
[133] K. Ekblom, J. Hultdin, B. Stegmayr, I. Johansson, B. Van Guelpen, G. Hallmans,
L. Weinehall, L. Johansson, P.G. Wiklund, S.L. Marklund, Iron stores and HFE
genotypes are not related to increased risk of ischemic stroke. A prospective nested
case-referent study, Cerebrovasc Dis, 24 (2007) 405-411.
[134] Y. Deugnier, J. Mosser, Modifying factors of the HFE hemochromatosis
phenotype, Expert Rev Gastroenterol Hepatol, 2 (2008) 531-540.
[135] K. Hashimoto, M. Hirai, Y. Kurosawa, Identification of a mouse homolog for the
human hereditary haemochromatosis candidate gene, Biochem Biophys Res
Commun, 230 (1997) 35-39.
[136] X.Y. Zhou, S. Tomatsu, R.E. Fleming, S. Parkkila, A. Waheed, J. Jiang, Y. Fei,
E.M. Brunt, D.A. Ruddy, C.E. Prass, R.C. Schatzman, R. O'Neill, R.S. Britton, B.R.
Bacon, W.S. Sly, HFE gene knockout produces mouse model of hereditary
hemochromatosis, Proc Natl Acad Sci U S A, 95 (1998) 2492-2497.
[137] M.S. Golub, S.L. Germann, R.S. Araiza, J.R. Reader, S.M. Griffey, K.C. Lloyd,
Movement disorders in the Hfe knockout mouse, Nutr Neurosci, 8 (2005) 239-244. 54
[138] D. Johnstone, R.M. Graham, D. Trinder, R.D. Delima, C. Riveros, J.K. Olynyk,
R.J. Scott, P. Moscato, E.A. Milward, Brain transcriptome perturbations in the Hfe(-/-
) mouse model of genetic iron loading, Brain research, 1448 (2012) 144-152.
[139] S. Tomatsu, K.O. Orii, R.E. Fleming, C.C. Holden, A. Waheed, R.S. Britton, M.A.
Gutierrez, S. Velez-Castrillon, B.R. Bacon, W.S. Sly, Contribution of the H63D
mutation in HFE to murine hereditary hemochromatosis, Proc Natl Acad Sci U S A,
100 (2003) 15788-15793.
55
Chapter 2
H63D HFE gene variant alters brain iron profiles and increases oxidative stress
2.1. Abstract
Because of the increasing evidence that H63D HFE polymorphism appears in higher frequency in neurodegenerative diseases, we evaluated the neurological consequences of H63D HFE in vivo using mice that carry H67D HFE (homologous to human H63D). Although total brain iron concentration did not change significantly in the
H67D mice, the expression of brain iron management proteins was altered significantly.
The 6-month-old H67D mice had increased HFE and H-ferritin expression. At 12-months
H67D mice had increased H- and L-ferritin but decreased transferrin expression suggesting increased iron storage and decreased iron mobilization. Increased L-ferritin positive microglia in H67D mice suggests that microglia increase iron storage to maintain brain iron homeostasis. The 6-month-old H67D mice had increased levels of GFAP, increased oxidatively modified protein levels, and increased cystine/glutamate antiporter
(xCT) and hemeoxygenase-1 (HO-1) expression indicating increased metabolic and oxidative stress. By 12 months there was no longer increased astrogliosis or oxidative stress. The decrease in oxidative stress at 12 months could be related to an adaptive response by nuclear factor E2-related factor 2 (Nrf2) that regulates antioxidant enzymes expression and is increased in the H67D mice. These findings demonstrate that the H63D
HFE impacts brain iron homeostasis, and promotes an environment of oxidative stress and induction of adaptive mechanisms. These data, along with literature reports on humans with HFE mutations provide the evidence to overturn the traditional paradigm 56
that the brain is protected from HFE mutations. The H67D knock-in mouse can be used as a model to evaluate how the H63D HFE mutation contributes to neurodegenerative diseases.
2.2. Introduction
Although the penetrance of H63D HFE for hereditary hemochromatosis (HH) is lower than the C282Y HFE, H63D HFE is associated with increased serum transferrin saturation, increased serum ferritin level and increased serum iron level, particularly in elderly populations (>55 years) and in some of the demographic subgroups such as
Mexican-American [1-3]. This biochemical penetrance of the H63D allele resulted in our investigations into the relationship between the H63D HFE genotype and late onset neurodegenerative diseases [4] where increased brain iron is often reported [5, 6].
The HFE protein is expressed in endothelial cells, choroid plexus and the ependymal cells where it can influence brain iron content [7]. However, based on misinterpretation of studies in the mid-1900s, the brain was thought to be protected from iron overload associated with HFE mutations. However, these studies in the mid-1900s
[8, 9] and recent MRI studies [10-12] all demonstrate an iron accumulation in the brain of
HH-patients; including those areas protected by the blood-brain-barrier. Moreover, in healthy aged individuals the presence of H63D HFE and transferrin C2 is associated with higher brain ferritin iron [13]. Increased hippocampal and basal ganglia iron is associated with poor declarative and verbal working memory [14]. However, Jahanshad et al. [15] recently reported a positive association between H63D HFE and white matter fiber integrity in healthy adults. 57
A relationship between iron accumulation and neurodegenerative disease is established [5, 6] so it was logical to consider a relationship between HFE genotypes and neurodegenerative diseases. The investigation into a relationship between HFE genotypes and neurodegenerative diseases was explored in the introduction to my thesis. The results of the population studies have been mixed and strongly suggest a gene environment interaction.
Thus, to determine the specific effects of H63D HFE on neurological phenotypes independent of environmental conditions we have developed cell models [16] and, in this paper, introduce an animal model, which carries an analogous HFE mutation (H67D
HFE). In a human neuroblastoma cell line expressing different HFE genotypes, the H63D cells have increased iron, oxidative stress [16] and endoplasmic reticulum stress [17], increased glutamate release and monocyte chemoattractant protein-1 secretion [18, 19] and tau phosphorylation [20]; each of which is proposed as a contributing factor to neurodegenerative diseases. Thus, we hypothesized that H63D HFE enables a convergence of mechanisms that promote pathogenic processes.
In this study we generated in vivo model, a H67D knock-in mouse line (mouse homologue of the human H63D), to evaluate the neurological consequences of H63D
HFE in vivo under controlled environmental conditions. We demonstrated that H63D
HFE alters brain iron homeostasis and creates an environment of oxidative stress. In the long-term H67D mice will serve as a model to explore how H63D HFE impacts disease mechanisms and therapeutic interventions in neurodegenerative disorders.
58
2.3. Material and methods
2.3.1. Generation of H67D Knock-in Mice
The H67D knock-in mice (mouse homologous to H63D in humans) were commercially generated (inGenious Targeting Laboratory, Inc, NY) as previously described by Tomatsu et al. [21]. Briefly, the HFE gene isolated from 129/SvJ mouse bacterial artificial chromosome library was subcloned into the pBS vector. The H67D point mutation (199C to –G) was introduced into exon 2 of HFE gene by site-directed mutagenesis, which destroyed a BspHI restriction site. The HFE gene fragment containing H67D mutation was added between the thymidine kinase (TK) and neor gene of a targeting vector (pPNT-loxP2 vector). The resulting targeting vector was linearized with NotI and introduced into the 129/Sv-derived embryonic stem (ES) cell line RW4
(Incyte Genomics Systems, St. Louis) by electroporation. ES clones that were resistant to both 200µg/ml G418 (BIBCO/BRL) and 2µM ganciclovir (Syntex Chemicals, Boulder,
CO) were isolated and used for injection into C57BL/6J blastocysts. Chimeric males were bred to C57BL/6J females for germ-line transmission. The F1 heterozygous mice were bred to Cre mice to remove neor gene flanked by loxP sites. The resultant neor- excised heterozygotes were interbred to generate wild-type (+/+), heterozygous (+/H67D) and homozygous (H67D/H67D) H67D knock-in mice.
Mice were maintained under normal housing conditions. They were given ad libitum access to rodent chow pellets and water. Both males and females were included in all experiments. All procedures were conducted according to the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Pennsylvania State University
College of Medicine Institutional Animal Care and Use Committee. 59
2.3.2. Mice Genotyping
DNA was isolated from tail biopsies according to DNeasy blood and tissue kit
(QIAGEN, CA). To amplify HFE gene including the H67D variant PCR analysis was conducted using a forward primer
(5’AGGACTCACTCTCTGGCAGCAGGAGGTAACCA 3’) and a reverse primer
(5’TTTCTTTTACAAAGCTATATCCCCAGGGT3’). PCR conditions were: 94ºC for 15 minutes, 94ºC for 45 seconds, 58ºC for 45 seconds, 39 cycles of 72ºC for 90 seconds and
72ºC for 10 minutes. Amplified PCR fragments were digested with BspHI restriction enzyme for 2 hours at 37ºC to detect H67D point mutation. DNA fragments were separated by 2% agarose gel electrophoresis.
2.3.3. Measurement of Iron
Brain and Liver samples were harvested from 6- and 12-month-old wild-type
(+/+), heterozygous (+/H67D) and homozygous (H67D/H67D) H67D knock-in mice (n =
4-6 per genotype). Samples were diluted 1:10 (wt:v) with 0.32 M sucrose and homogenized. Total brain and hepatic iron concentrations (µg/g of tissue; wet weight) were measured triplicate by graphite furnace atomic absorption spectrometry (model
5100AA, Perkin-Elmer, Norwalk, CT) according to standard protocol [22].
2.3.4. Immunoblotting
Brain and liver tissues from 6- and 12-month-old wild-type (+/+), +/H67D and
H67D/H67D mice (n = 4-6 per genotype) were homogenized in homogenization buffer:
1xPBS, 0.5% NP-40 (IGEPAL; Sigma, St. Louis, MO) and protease inhibitor cocktail 60
(1:100; Sigma, St. Louis, MO). The total protein concentration was determined with
Pierce BCA protein assay kit (Thermo scientific, MA). Total brain or liver homogenates
(20 µg of protein) were separated by electrophoresis in Criterion polyacrylamide Tris-
HCl gel (4-20%; Bio-Rad, Hercules, CA). Proteins were then transferred to nitrocellulose membranes and the membranes were blocked with 5% nonfat dry milk for one hour at room temperature. After overnight incubation at 4ºC with primary antibodies, membranes were then incubated with enhanced chemiluminescent (ECL) anti-host horseradish peroxidase-linked secondary antibodies (Amersham Bioscience, Piscataway, NJ) for an hour at room temperature. The signal was visualized by ECL detection (Perkin Elmer,
Waltham, MA) and Multigauge software (V3.0; Fuji film system) was used to quantitate the intensity of the band. Following primary antibodies were used: HFE (1:500; Sigma,
St. Louis, MO), H-ferritin (1:1000; Covance, Princeton, NJ), L-ferrtin (1:500; abcam,
Cambridge, MA), transferrin receptor (1:500; ; Zymed Laboratories Inc., San Francisco,
CA), divalent metal transporter-1 (1:1000; Covance, Princeton, NJ), transferrin (1:1000;
MP biomedicals, Solon, OH), T cell immunoglobulin and mucin domain-containing protein-2 (Tim-2; 1:2000; abcam, Cambridge, MA), cystine/glutamate antiporter (xCT;
1:500; abcam, Cambridge, MA), hemeoxygenase-1 (HO-1; 1:500; Enzo Life Science,
Farmingdale, NY), Nrf2 (1:1000; abcam, Cambridge, MA), GFAP (1:2000; Dako,
Carpinteria, CA) and beta-actin (1:3000; Sigma, St. Louis, MO).
2.3.5. Histology
Six- and 12-month-old wild-type (+/+) and H67D/H67D mice (n = 4 per genotype) were perfused transcardially with Ringer’s solution followed by ice-cold 4% 61
paraformaldehyde. The brains were paraffin-embedded and sectioned coronally at 6-µm- thick. The sections were deparaffinized and then rehydrated through a series of ethanol.
After antigen retrieval with sodium citrate (pH 6), endogenous peroxidase activity was blocked with hydrogen peroxide (3.7% in methanol) for 20 minutes at room temperature.
The sections were then blocked for one hour with 2% milk and were incubated with primary antibodies overnight at 4ºC followed by an hour incubation at room temperature with biotinylated anti-host secondary antibody (1:200; vector laboratories, Burlingame,
CA). Immunoreactivity was detected using the avidin biotin complex (ABC) and 3,3’- diaminobenzidine (DAB; vector laboratories, Burlingame, CA). The sections were analyzed with a bright-field microscopy by an investigator blinded for genotypes.
Following primary antibodies were used for immunostaining: L-ferritin (1:250; abcam,
Cambridge, MA), transferrin receptor (1:250; Zymed Laboratories Inc., San Francisco,
CA), divalent metal transporter-1 (1:200; convence, Princeton, NJ) and transferrin
(1:200; MP biomedicals, Solon, OH).
For immunofluorescence, after overnight incubation with rabbit anti-GFAP antibody (1:1000; Dako, Carpinteria, CA) or rabbit Iba-1 antibody (1:600; Wako,
Richmond, VA), sections were probed with Alexa Flour 488 secondary antibody
(Invitrogen, Grand Island, NY) for 1 hour in the dark at room temperature. After washes, slides were mounted and the sections were analyzed with fluorescence microscopy. For double immunofluorescence staining, because both L-ferritin and Iba-1 antibodies were made from the same host we fluorescently labeled L-ferritin using a DyLight 550 antibody labeling kit (Thermo fisher scientific, Waltham, MA) prior to the incubation with brain sections. The sections were first incubated overnight with rabbit Iba-1 62
antibody followed by overnight incubation with DyLight 550 labeled L-ferritin. The sections were then probed with Alexa Flour 488 secondary antibody for 1 hour in the dark at room temperature. After washes, the slides were mounted and the sections were analyzed with fluorescence microscopy.
2.3.6. Myelin Isolation and Analysis of Myelin Protein
Brain samples from 6- and 12-month-old mice (5-6 per genotype) were weighed and homogenized with 1.5 ml of 0.32M sucrose. After adding 2 ml of 0.85M sucrose samples were centrifuged for 30 minutes at 41000 rpm at 4ºC. The supernatant was discarded; the middle layer was collected and resuspended with 2.5 ml iron free water
(Sigma, St. Louis, MO). The samples were centrifuged for 15 min at 41000 rpm at 4ºC.
Supernatant was discarded and the pellet was resuspended with 3.0 ml iron free water.
The samples were centrifuged 10 min at 17000 rpm at 4ºC and collected the pellet, which was crude myelin protein. The pellets were kept at -80ºC for 20 min and then were freeze-dry using a lyophilizer overnight. The pellets were then left at room temperature for 20 minutes and were weighed. The pellets were resuspended with RIPA buffer
(Sigma, St. Louis, MO) and centrifuged 45 min at 51,000 rpm at 4 ºC. Supernatant was collected and protein concentration was determined with Pierce BCA protein assay kit
(Thermo scientific, MA). Total protein of 10 µg was used for immunoblot analyses to determine the expression of myelin basic protein (MBP; 1:1000; abcam, Cambridge,
MA), proteolipid protein (PLP; 1:1000; Millipore, Billerica, MA) and CNPase (1:500; abcam, Cambridge, MA).
63
2.3.7. Measurement of Oxidatively Modified Proteins
As a consequence of oxidative modification to proteins, carbonyl groups are introduced to the side chain of amino acids. These carbonyl groups hallmark the oxidative status of protein [23]. An Oxyblot kit (Millipore, Billerica, MA) was used to measure protein carbonyl levels. Briefly, total brain homogenates (20 µg protein) from 6- and 12-month-old wild-type, +/H67D and H67D/H67D mice (6 per genotype) were reacted with 2,4-dinitrophenylhydrazone (DNP-hydrazone). Samples were then treated according to the manufacturer’s protocol.
2.3.8. Statistical Analyses
Data were expressed as mean ± standard error. An analysis of variance (one-way
ANOVA; GraphPad Prism 4) followed by a Tukey multiple comparison or Dunnett test was used to compare between experimental groups. For iron measurement, two-way
ANOVA was performed to analyze the interaction of age and genotypes with iron concentration. Bonferroni posttest was used to compare between experimental parameters. A value of p< 0.05 was considered significant for all experiments.
2.4. Results
2.4.1. Generation of H67D Mice
A H67D knock-in mouse line (mouse homologue of the human H63D) was generated by site-directed mutagenesis to murine HFE gene. The resultant heterozygous mice were bred to generate wild-type, heterozygous and homozygous H67D knock-in mice. Mice carrying the H67D variant were indistinguishable at birth from their 64
littermate controls. Homozygous H67D mice develop normally and are reproductively viable. Genotyping of mice was performed by PCR analysis of DNA obtained from tail biospies and subsequent digestion with BspHI restriction enzyme (Figure 2-1 A). Body weights of mice were taken at 6- and 12-months of age. At both ages, H67D/H67D mice had significantly higher body weight than wild-type mice while the heterozygous mice
(+/H67D) had similar body weight as the wild-type (Figure 2-1 B).
To confirm whether the allelic variant is functional we determined hepatic iron levels in 6- and 12-month-old H67D mice. Compared to the wild-type mice, 6- and 12- month-old H67D/H67D mice had a 67% (266.4 ± 27.3 vs. 159.6 ± 31.9 µg/g of liver) and
70% (281.35 ± 50.7 vs. 165.7 ± 22.9 µg/g of liver) increase in hepatic iron concentration while +/H67D mice had similar hepatic iron concentration as wild-type mice at both ages
(Figure 2-1 C).
65
Figure 2-1
66
Figure 2-1. Increased body weight and hepatic iron concentration in H67D knock-in mice.
(A) Generation of H67D knock-in mice. PCR analysis of DNA obtained from tail
biopsies was followed by digestion with BspHI restriction enzyme to detect the
H67D point mutation. DNA digested with BspHI from wild-type mice (+/+) results
240 and 260 bp; DNA digested with BspHI from +/H67D mice results 500, 240 and
260 bp and DNA digested with BspHI from H67D/H67D results 500 bp.
(B) Body weight was taken at 6 and 12 months of age in three groups. At both ages, body
weight of H67D/H67D mice is higher compared to the wild-type mice. Bars
represent mean ± standard error. (* = p < 0.05, ** = p < 0.01; n = 14 to 36 per
genotype for each age group).
(C) Total iron concentration measured by atomic absorption spectrometry is 67% and
70% increase in the liver of 6- and 12-month-old H67D/H67D mice compared to the
wild-type (+/+) mice. Hepatic iron concentration in H67D/H67D mice is 65% and
64% increase compared to +/H67D mice for both age groups. Bars represent mean ±
standard error. * represents a significant difference from wild-type (p < 0.05) and #
represents a significant difference from +/H67D (p < 0.05). n = 4 to 9 per genotype.
67
Six and 12-month-old H67D/H67D mice also had changes in the expression of hepatic iron management proteins. Consistent with increased iron concentration, iron storage proteins, H-ferritin and L-ferritin were significantly increased while iron transport protein transferrin receptor (TfR) was significantly decreased in H67D/H67D mice compared to wild-type mice at both ages. Ceruloplasmin, which is involved in iron export, was also significantly decreased in 12-month-old H67D/H67D mice compared to the wild-type (Figure 2-2). Together these results indicated that the allelic variant is functional in H67D mice.
68
Figure 2-2
69
Figure 2-2. Altered hepatic iron management protein expression in H67D knock-in mice.
Liver homogenates from 6- and 12-month-old wild-type (+/+), +/H67D and H67D/H67D were determined for the presence of H-ferritin, L-ferritin, transferrin receptor, and cerruloplasmin by Western blot. A representative Western blot is shown for each protein and quantification of blots is shown as bar graphs. The expression level was normalized to β-actin. Bars represent mean ± standard error. * represents a significant difference from wild-type (* = p < 0.05; ** = p < 0.01; *** = p < 0.001). # represents a significant difference from +/H67D (# = p < 0.05; ## = p < 0.01; n = 4 to 6 per genotype).
(A-D) At 6 months of age, H-ferritin (A) and L-ferritin (B) expressions are increased
while transferrin receptor expression is decreased (C) in H67D/H67D mice.
Ceruloplasmin (D) expression does not change significantly in H67D knock-in
mice.
(E-H) At 12 months of age, H-ferritin (E) and L-ferritin (F) expressions are increased
while transferrin receptor (G) and ceruloplasmin (H) expressions are decreased in
H67D/H67D compared to +/+ mice.
70
2.4.2. Brain Iron Concentration in H67D Mice
Total brain iron was measured in 6- and 12-month-old mice by atomic absorption spectrometry. At 6 months, total brain iron level was 22% higher in H67D/H67D (22.57
± 2.0 µg/g of brain) and 18.6% higher in +/H67D mice (21.95 ± 1.4 µg/g of brain) compared to wild-type mice (18.5 ± 0.7 µg/g of brain) although these differences did not reach statistical significance (p = 0.26). At 12 months of age, total brain iron concentration in H67D/H67D mice (24.04 ± 1.1 µg/g) and +/H67D mice (27.76 ± 3.0
µg/g) was not different compared to wild-type mice (24.58 ± 1.4 µg/g). There was a significant interaction for age and iron concentrations indicating that regardless of genotypes brain iron concentration increased with age (p = 0.01). Compared to 6-month- old mice, by 12 months, brain iron concentration increased by 33% in the wild-type,
26.5% in the +/H67D but only 6.5% in H67D/H67D mice (Figure 2-3 and Table 2-1).
71
Figure 2-3
Figure 2-3. Total brain iron concentration is altered in H67D knock-in mice. Total brain iron was measured with atomic absorption spectrometry in 6- and 12-month old mice.
Brain iron concentration is 22% and 18.6% increase in H67D/H67D and +/H67D compared to wild-type (+/+) mice at 6 months (p = 0.26). At 12 months of age brain iron level in H67D knock-in mice is not significantly different from the wild-type mice.
Analysis of brain iron level by age indicates wild-type and +/H67D mice have increased brain iron with age (32.9% and 26.5% respectively). Brain iron level is increased only
6.5% in 12-month-old H67D/H67D mice compared to 6-month-old H67D/H67D mice.
Bars represent mean ± standard error. (n = 4 to 8 per genotype in both age groups).
72
2.4.3. Alteration in Expression of Proteins Involved in Iron Homeostasis in H67D Mice
Although total iron levels are important, it is the response of the regulatory proteins that are perhaps most important because they maintain homeostasis. Thus, we evaluated the expressions of proteins involved in brain iron homeostasis: HFE, H-ferritin,
L-ferritin, transferrin receptor (TfR), transferrin (Tf), divalent metal transporter-1 (DMT-
1) and T cell immunoglobulin and mucin domain-containing protein-2 (Tim-2). Western blot analyses of brain homogenates from 6- and 12-month-old mice revealed differences in the levels of iron management proteins in H67D/H67D mice compared to the wild- type. At 6 months H67D/H67D mice had significant increases in HFE (Figure 2-4 A) and
H-ferritin (Figure 2-4 B) and a 40% decrease in DMT-1 which did not reach statistical significance (p = 0.07; Figure 2-4 D). L-ferritin, Tf, Tim-2 levels (Figure 2-4 C, E, F) and
TfR (data not shown) in H67D/H67D mice were not different from wild-type mice.
None of the iron management protein levels in +/H67D mice was different from wild- type at 6 months of age (Figure 2-4 A-F).
73
Figure 2-4
74
Figure 2-4. Altered brain iron management protein expression in 6-month-old H67D knock-in mice. Brain homogenates from 6-month-old wild-type (+/+), +/H67D and
H67D/H67D were determined for the presence of HFE, H-ferritin, L-ferritin, transferrin, divalent metal transporter-1 and T cell immunoglobulin and mucin domain-containing protein-2 (Tim-2) by Western blot. A representative Western blot is shown for each protein and quantification of blots is shown as bar graphs. The expression level was normalized to β-actin. At 6 months of age, HFE (A) and H-ferritin (B) expression is increased while DMT-1 expression is tended to decrease (D; p = 0.07) in H67D/H67D mice. L-ferritin (C), transferrin (E) and Tim-2 expressions (F) are not significantly changed in H67D knock-in mice. Bars represent mean ± standard error. * represents a significant difference from wild-type (* = p < 0.05; ** = p < 0.01). # represents a significant difference from +/H67D (### = p < 0.001; n = 4 to 6 per genotype).
75
At 12 months of age, H67D/H67D mice had increases in H- and L-ferritin levels
(Figure 2-5 B, C), and decreases in Tf (Figure 2-5 E) and Tim-2 levels (Figure 2-5 F) compared to wild-type mice. The relative concentrations of HFE, DMT-1 (Figure 2-5 A,
D) and TfR (not shown) in H67D/H67D mice were not different from wild-type mice at this age. None of the iron management protein levels in +/H67D mice was different from wild-type at 12 months (Figure 2-5 A-F).
76
Figure 2-5
77
Figure 2-5. Altered brain iron management protein expression in 12-month-old H67D knock-in mice. Brain homogenates from 12-month-old wild-type (+/+), +/H67D and
H67D/H67D were determined for the presence of HFE, H-ferritin, L-ferritin, transferrin, divalent metal transporter-1 and T cell immunoglobulin and mucin domain-containing protein-2 (Tim-2) by Western blot. A representative Western blot is shown for each protein and quantification of blots is shown as bar graphs. The expression level was normalized to β-actin. At 12 months of age, H-ferritin (B) and L-ferritin (C) expressions are increased while transferrin (E) and Tim-2 expressions (F) are decreased in
H67D/H67D compared to +/+ mice. HFE (A) and DMT-1 expressions (D) in
H67D/H67D mice do not alter significantly. Bars represent mean ± standard error. * represents a significant difference from wild-type (* = p < 0.05; ** = p < 0.01). # represents a significant difference from +/H67D (# = p < 0.05; ## = p < 0.01; n = 4 to 6 per genotype).
78
We also determined the distribution and cellular localization of TfR, DMT-1, transferrin and L-ferritin by immunohistochemical staining in both age groups. At the cellular level, tranferrin receptor (TfR) staining was primarily confined to neurons with very little or no TfR staining in the white matter tracts in either group. There appeared to be regional differences in staining for TfR at 6 months. Compared to wild-type mice,
H67D/H67D mice had weaker neuronal staining for TfR in the cortex (Figure 2-6 A-B) and the cerebellum (Figure 2-6 E-F) but stronger TfR staining in the striatum (Figure 2-6
C-D).
79
Figure 2-6
80
Figure 2-6: Immunohistochemical localization for transferrin receptor (TfR). Neurons in the cortex (A, B) and Purkinje cells in the cerebellum (arrows in E, F) of both wild-type
(+/+) and H67D/H67D mice express TfR; however TfR staining is weaker in
H67D/H67D mice. In striatum, TfR staining is primarily confined to neurons in both groups with little or no TfR staining is observed in the white matter tract. Compared to wild-type, H67D/H67D mice have relatively higher staining intensity for TfR in the striatum (C, D). A scale bar represents 50 µm. n = 4 per genotype.
81
Immunohistochemical analysis for DMT-1 revealed that similar cell types were stained for DMT-1 protein in both groups (Figure 2-7). However, compared to wild-type mice (Figure 2-7 A, C, E), 6-month-old H67D/H67D mice had relatively weaker neuronal staining for DMT-1 in the cortex (Figure 2-7 B) and the purkinje cells of cerebellum (Figure 2-7 F), and weaker staining in oligodendrocytes in the striatum
(Figure 2-7 D).
82
Figure 2-7
83
Figure 2-7: Immunohistochemical localization for divalent metal transporter-1(DMT-1).
Neurons in the cortex of both wild-type (+/+) and H67D/H67D mice express DMT-1 with much less DMT-1 staining is observed in H67D/H67D mice compared to wild-type mice (arrows in A, B). In striatum, DMT-1 expresses predominantly in oligodendrocytes
(arrows in C, D). Blood vessels staining for DMT-1 are also observed in both wild-type and H67D/H67D mice; however, the staining intensity is relatively weaker in H67D mice
(arrow head in C, D). In cerebellum, strong neuronal staining for DMT-1 is observed in purkinje cells in wild-type mice (arrows in E). In contrast, purkinje cells of cerebellum in
H67D/H67D mice have relatively little staining for DMT-1 (arrows in F). bv = blood vessel. A scale bar represents 50 µm. n = 4 per genotype.
84
For transferrin, the same type of cells was positive for staining in both wild-type and H67D/H67D mice. Neuronal transferrin staining was observed in the cortex (Figure
2-8 A, B) while transferrin staining was found predominantly in oligodendrocytes in the striatum and corpus callosum (Figure 2-8 C-F). Staining intensity for transferrin was much less in H67D/H67D (Figure 2-8 B, D, F) compared to wild-type mice in all regions examined at 12 months (Figure 2-8 A, C, E).
85
Figure 2-8
86
Figure 2-8: Immunohistochemical localization for transferrin. Neurons in the cortex of both wild-type (+/+) and H67D/H67D mice express transferrin (arrows in A, B). Few transferrin positive oligodendrocytes are observed in the cortex (arrow head in A, B).
However, transferrin expresses predominantly in oligodendrocytes in the straitum (arrows in C, D) and corpus collosum (arrows in E, F) of both wild-type and H67D/H67D mice.
Transferrin positive oligodendrocytes appear in a row in the corpus collosum (arrows in
E, F) and only a few of process bearing oligodendrocytes staining for transferrrin are present (arrow heads in E, F). In all observed regions, transferrin staining is less in
H67D/H67D mice compared to wild-type mice. LV = lateral ventricle; bv = blood vessel.
A scale bar represents 50 µm. n = 4 per genotype.
87
Higher L-ferritin staining intensity was seen only in 12-month-old H67D/H67D mice compared to wild-type mice, which is consistent with immunoblot analysis. L- ferritin staining was found primarily in the glial cells, although there was relatively weak neuronal staining in the cortex of H67D/H67D (Figure 2-9 A-B). L-ferritin staining was primarily found in the oligodendrocytes in the striatum and thalamus (Figure 2-9 C-F) in both groups; however, there was more robust L-ferritin staining in H67D/H67D mice
(Figure 2-9 B, D, F) compared to the wild-type mice (Figure 2-9 A, C, E).
88
Figure 2-9
89
Figure 2-9: Immunohistochemical localization for L-ferritin. L-ferritin staining in the cortex is relatively weak in both wild-type (+/+) and H67D/H67D mice. The staining is mostly confined to glial cells (arrows in B). L-ferritin staining in the striatum (arrows in
C, D) and the thalamus (E, F) is primarily confined to oligodendrocytes. L-ferritin staining in the cortex, striatum and the thalamus is relatively weak in wild-type mice (A,
C, E). In contrast L-ferritin staining is more robust with more L-ferritin-positive oligodendrocytes in H67D/H67D mice (B, D, F). A scale bar represents 50 µm. n = 4 per genotype.
90
To determine whether L-ferritin positive glial cells observed in the cortex were either astrocytes or microglia, we performed double immunofluorescence staining. The fluorescence analysis also reveals an increase L-ferritin staining in 12-month-old
H67D/H67D mice and strong co-localization of L-ferritin with Iba-1 positive microglia
(Figure 2-10). There also appeared to be an increase in number of Iba-1 positive microglia and L-ferritin immunoreactive microglia in 12-month-old H67D/H67D indicating the presence of microgliosis in H67D mice (Figure 2-10 E, F). However, the number of microglia in H67D mice at 6-months did not differ from wild-type mice
(Figure 2-11). L-ferritin did not co-localize with GFAP positive astrocytes (Figure 2-12).
91
Figure 2-10
92
Figure 2-10: Double immunostaining showing the co-localization of L-ferritin and Iba-1.
Double fluorescence analysis indicates that L-ferritin fluorescence staining is present primarily in the glial cells in the cortex of both wild-type (+/+) and H67D mice (A, D); however there are increased number of L-ferritin positive glial cells in H67D mice (D).
Merged images indicate the co-localization of L-ferritin with Iba-1 positive microglia to greater extent (C, F). More L-ferritin positive microglia is observed in the cortex of
H67D mice (E, F) compared to the wild-type (B, C). A scale bar represents 50 µm. n = 4 per genotype.
93
Figure 2-11
Figure 2-11: Immunofluorescence staining for Iba-1. Brain sections from 6-month-old mice were stained for Iba-1 to detect microglia. Immunofluorescence analysis indicates that number of microglia in the cortex is not different between wild-type (+/+) and H67D mice. A scale bar represents 50 µm. n = 4 per genotype.
94
Figure 2-12
95
Figure 2-12: Double immunostaining showing no co-localization of L-ferritin and GFAP.
Double fluorescence analysis indicates that relatively higher L-ferritin fluorescence staining is present in the cortex of H67D mice (D) compared to the wild-type (A).
Merged images indicate that L-ferritin and GFAP positive astrocytes are not co-localized
(C, F). A scale bar represents 50 µm. n = 4 per genotype.
96
2.4.4. Myelin Proteins Were Not Changed in H67D Mice
Crude myelin protein was extracted by sucrose gradient and ultracentrifugation.
Three major myelin proteins in myelin extract were analyzed in 6- and 12-month-old mice by immunoblotting. Myelin basic protein (MBP), proteolipid protein (PLP) and
CNPase protein expression were not changed significantly between any of the groups at either age (Figure 2-13 A-F and Table 2-1).
97
Figure 2-13
98
Figure 2-13: Myelin proteins are not altered in H67D knock-in mice. Expression of three major myelin proteins in myelin extract from brains of 6- and 12-month-old mice were analyzed by Western blot. A representative Western blot is shown for each protein and quantification of blots is shown as bar graphs. The expression level was normalized to β- actin. There is no difference in expressions of myelin basic protein (A, D), CNPase (B, E) and proteolipid protein (C, F) between three groups at both ages. Bars represent mean ± standard error. (n = 5 to 6 per genotype in both age groups).
99
2.4.5. H67D HFE is Associated with Astrogliosis
To determine whether H67D HFE is associated with increased gliosis, we evaluated GFAP staining in H67D mice at both ages (Figure 2-14, Figure 2-15 and Table
2-1). GFAP staining was increased in all observed brain regions; the cortex, hippocampus and the cerebellum in 6-month-old H67D/H67D (Figure 2-14 C, E, G, I ) compared to wild-type mice (Figure 2-14 B, D, F, H), which is consistent with the significant increase in total GFAP level in H67D knock-in mice determined by immunoblotting (Figure 2-14
A). GFAP staining and total GFAP level in 12-month-old H67D mice was not changed significantly from wild-type mice (Figure 2-15 A-I).
100
Figure 2-14
101
Figure 2-14: Increased GFAP in 6-month-old H67D knock-in mice.
(A) GFAP protein expression was determined in brain homogenates from 6-month-old
mice. Quantification of blots is shown as bar graphs. The expression level was
normalized to β-actin. At 6 months, the +/H67D and H67D/H67D mice have
significant increase in GFAP protein expression compared to +/+ mice. Bars
represent mean ± standard error. (* = p < 0.05; ** = p < 0.01; n = 5 to 6 per
genotype).
(B-I) Immunofluorescence analyses of GFAP in brains of 6-month-old mice indicates
that GFAP immunoreactivity is increased in the cortex (C), hippocampus (E and G)
and cerebellum (I) of H67D/H67D compared to +/+ mice (B, D, F and H). A scale bar
represents 50 µm. (n = 4 per genotype).
102
Figure 2-15
103
Figure 2-15: GFAP levels are not increased in 12-month-old H67D knock-in mice.
(A) GFAP protein expression was determined in brain homogenates from 12-month-old
mice. Quantification of blots is shown as bar graphs. The expression level was
normalized to β-actin. Total brain GFAP expression in 12-month-old +/H67D and
H67D/H67D mice did not change significantly compared to +/+ mice. Bars represent
mean ± standard error. (n = 5 to 6 per genotype).
(B-I) Immunofluorescence analyses of GFAP in brains of 12-month-old mice indicates
that GFAP immunoreactivity in the cortex (C), hippocampus (E and G) and
cerebellum (I) of H67D/H67D mice is not different compared to +/+ mice (B, D, F
and H). A scale bar represents 50 µm. (n = 4 per genotype).
104
2.4.6. H67D HFE is Associated with Oxidative stress
Because H63D HFE is associated with higher baseline stress in our cell models
[16], we determined the level of oxidatively modified proteins (carbonyl level), and expression of xCT antiporter, which is essential for maintaining intracellular glutathione level [24], and hemeoxygenase-1 (HO-1; [25]) as indices of oxidative stress in H67D mice. The level of oxidatively modified proteins increased in both H67D/H67D (64%) and +/H67D (50%) mice at 6 months of age compared to the wild-type (Figure 2-16 A and Table 2-1). Moreover, xCT expression was 86% higher and HO-1 expression was
27% higher in the brains of 6-month-old H67D/H67D mice compared to the wild-type mice (Figure 2-16 C, E and Table 2-1). However, there was no difference in the oxidatively modified proteins level, and xCT and HO-1 expression between three groups at 12 months of age (Figure 2-16 B, D, F and Table 2-1).
105
Figure 2-16
106
Figure 2-16: Increased oxidative stress in H67D knock-in mice.
An oxyblot assay and slotblot analysis was used to measure total oxidatively modified protein levels in 6-month-old and 12-month-old mice.
(A) Total oxidatively modified protein levels is significantly higher in 6-month-old
+/H67D (50%) and H67D/H67D (64%) compared to wild-type (+/+) mice.
(B) By 12 months, total oxidatively modified protein levels in H67D mice do not change
significantly compared to +/+ mice.
Brain homogenates from 6- and 12-month-old wild-type (+/+), +/H67D and
H67D/H67D were determined for the expression of cystine/glutamate antiporter (xCT) that is essential for maintaining intracellular glutathione level and hemeoxygenase-1
(HO-1). A representative Western blot is shown and quantification of blots is shown as bar graphs. The expression level was normalized to β-actin. Bars represent mean ± standard error. (* = p < 0.05; ** = p < 0.01; *** = p < 0.001; n = 5 to 6 per genotype).
(C, E) The expression of xCT and HO-1 in 6-month-old H67D/H67D is increased by
86% and 27% respectively compared to the wild-type (+/+) mice.
(D, F) Brain xCT and HO-1 expression in H67D/H67D and +/H67D mice is not
significantly different compared to the wild-tpe (+/+) mice at 12-months.
107
To examine whether the increase in oxidative stress observed in 6-month-old
H67D mice, is associated with an impaired cellular defense system against redox stress we measured the expression of nuclear factor E2-related factor 2 (Nrf2) that controls endogenous anti-oxidant genes transcription. The expression of Nrf2 in 6-month-old
H67D/H67D and +/H67D mice was not different from wild-type mice (Figure 2-17 A and Table 2-1) but was 69% higher (p < 0.05) in 12-month-old H67D/H67D mice compared to the wild-type (Figure 2-17 B and Table 2-1).
108
Figure 2-17
Figure 2-17: H67D HFE increases Nrf2 expression. Brain homogenates from 6- and 12- month-old wild-type (+/+), +/H67D and H67D/H67D were determined for the expression of nuclear factor E2-related factor 2 (Nrf2) that controls the endogenous anti-oxidant genes transcription in response to redox stress. A representative Western blot is shown and quantification of blots is shown as bar graphs. The expression level was normalized to β-actin. Bars represent mean ± standard error. (* = p < 0.05; n = 5 to 6 per genotype in both age group).
(A) The expression of Nrf2 in 6-month-old H67D/H67D and +/H67D mice is not
different significantly from wild-type (+/+) mice.
(B) Brain Nrf2 expression level is increased in 12-month-old H67D/H67D mice
compared to the wild-type mice.
109
Table 2-1
Total expression level in the brain (relative to wild-type) Proteins 6-month-old H67D mice 12-month-old H67D mice
Iron 22% Not different
HFE Not different
H-ferritin
L-ferritin Not different
Transferrin receptor Not different Not different
Transferrin Not different
Divalent metal transporter-1 40% Not different
Tim-2 Not different
Myelin proteins Not different Not different
Iba-1 (microglia) Not different
GFAP Not different
Oxidatively modified proteins Not different Cystine/glutamate transporter Not different (xCT)
Hemeoxygenase-1 (HO-1) Not different
Nrf2 Not different
Table 2-1. Summary findings from neurological characterization of the H67D mice.
110
2.5. Discussion
Because liver iron accumulation is a common histologic finding and perhaps the gold standard for diagnosis of HFE associated hemochromatosis we first determined the liver iron concentration in H67D mice. It is notable that even on standard diet, total hepatic iron content was significantly elevated in both 6- and 12-month-old H67D/H67D mice, which is consistent with the previous report of Tomatsu et al. [21] for 10-week-old
H67D mice. H67D/H67D mice also had significant alterations in expression of liver iron management proteins such as increased storage proteins while decreased transport proteins. Changes in liver iron profiles in H67D mice indicate that the H67D allelic variant is functional in these mice. Although H67D mice are indistinguishable from wild- type mice at birth, H67D/H67D mice have increased weight gain with age. Increased body weight together with higher hepatic iron concentration in H67D mice may be relevant to the increased prevalence of metabolic syndrome including obesity that is positively associated with elevated iron stores and increased serum ferritin levels [26-29].
Thus, the utility of this model could extend beyond studies directly involving the nervous system and have relevance to studies on HFE mutations in general. Because HFE polymorphisms are common allelic variants in Caucasians, the H67D mouse line presented here is a meaningful model for human diseases.
This model differs from previous investigations that used HFE knockout mice to study the association between HFE and brain iron accumulation. Although both our model and the knockout model accumulate iron in the liver [21, 30], only our H67D knock-in model showed brain iron accumulation. However, the HFE knockout mice were examined at an earlier age than ours [31, 32]. Our model with an analogous HFE 111
mutation (H67D HFE) is more similar to the human condition and MRI studies [10-12] and early histological studies [8, 9] that report increased iron in the brain in association with HFE mutations. The presence of the HFE mutation is associated with multiple changes in cells and animal models, including ER stress [17] and cholesterol disruption
[33] which suggests the HFE mutant protein may impact the cell phenotype beyond iron uptake.
Our study evaluated brain iron profiles in H67D knock-in mice at 6 and 12 months of age. Brain iron was increased by 22% in H67D mice, but this was not statistically significant. However, there was increased expression of iron storage protein
H-ferritin at both ages as well as an increase in L-ferritin at 12-months in the H67D mice.
These findings are a strong indication that the increased amount of brain iron in H67D mice was biologically meaningful and suggest brain iron metabolism is altered in these mice. Perhaps the most impressive evidence for altered brain iron homeostasis in H67D mice is the gliosis and increased oxidative stress indicated by increased oxidatively modified proteins, and increased expression of xCT and HO-1. Of note, is that the consequences of loss of iron homeostasis (increased oxidative stress and astrogliosis) are manifested in the brains of the 6-month-old mice but not at 12 months. The lack of oxidative stress in 12 months of age could be the result of an apparent adaptive response to the oxidative stress mediated by Nrf2 which was elevated in the 12-month-old H67D mice.
In addition to the increase in Nrf2 as an adaptive response, there are multiple adaptive responses by the iron management proteins. At 6 months, H67D mice have increased H-ferritin which is involved in rapid iron uptake and iron detoxification [34, 112
35], and a regional decrease in TfR, which delivers iron to neurons [36, 37]. The expression of ferritin and TfR are regulated post-transcriptionally according to iron status. Iron overload increases ferritin protein synthesis while decreases TfR protein synthesis [6]. Thus, increased H-ferritin is consistent with increased iron concentration in
H67D mice. Although we expected a decrease in TfR expression, total TfR expression in
H67D mice was not different from wild-type mice. At the cellular level however, we found regional differences in TfR staining in H67D mice with relatively weaker in the cortex and cerebellum but increased TfR staining in the striatum compared to the wild- type. Regional differences in TfR expression has been reported by Dornelles et al. [38] who demonstrated that iron loading decreased TfR mRNA in the cortex and hippocampus while increased TfR mRNA in the striatum. Regional differences in TfR expression may be associated with heterogeneity of iron distribution in the brain [39, 40]. In all brain regions examined, we found that TfR staining was primarily confined to neurons in both groups which are consistent with previous studies [36, 37, 41].
Transferrin, the major iron mobilization protein, is 39% decreased in H67D mice.
Because transferrin is synthesized primarily by the oligodendrocytes [42] reduced transferrin may suggest decreased metabolic activity by oligodendrocytes. However, transferrin is also taken up by the brain [43, 44] and iron overload lowers transferrin uptake by the brain [43]. Decreased transferrin levels are also associated with aging and brain disease [45]. Together these data suggest lower transferrin levels in H67D mice would decrease iron mobilization and limit iron delivery, particularly to neurons. The other iron delivery protein in the brain, Tim-2 is decreased in 12-month-old H67D mice.
Tim-2 is a receptor for H-ferritin on oligodendrocytes and its expression is inversely 113
related to iron status [46]. Therefore, decreased Tim-2 expression in H67D mice is an additional response to decrease iron availability. Despite the decrease in transferrin and
Tim-2, myelin protein expression is normal but total brain cholesterol is decreased [33].
Cholesterol is expressed by neurons as well as found in myelin, thus further analysis of myelin in the presence of the HFE gene variants is warranted particularly in light of the diffusion weighted imaging (DWI) analysis by Jahanshad et al. [15] suggesting that
H63D HFE is associated with increased myelin integrity.
An additional compensatory response indicating increased iron availability in the brain is the 40% decrease in DMT-1 levels in 6-month-old H67D mice. DMT-1 transports iron out of the endosome to the cytosol of the endothelial cell and is down- regulated with increased iron availability. DMT-1 is found in brain endothelial and ependymal cells, and a mutation in DMT-1 led to less detectable iron in both neurons and glia [47, 48]. In H67D mice, DMT-1 staining is lower in the cortex, striatum and cerebellum suggesting that cells are exposed to increased iron. Further evidence for altered brain iron homeostasis in 12-month-old H67D mice is a 2-fold increase in L- ferritin, which involves in long-term iron storage [49]. Increased L-ferritin staining is found in microglia and neurons; the latter being visible but weak. Although L-ferritin is primarily found in glial cells [50], neurons have capacity to express L-ferritin when challenged with iron [51]. Therefore, the presence of L-ferritin positive microglia and neurons in H67D mice suggests the initiation of adaptive responses and that microglia increases iron storage as an attempt to effectively manage the iron that has accumulated in the brain. 114
An additional change in glia in the H67D mice is the increased GFAP expression throughout the brain at 6 months indicating astrogliosis. Reactive gliosis is a rapid response to CNS injury and metabolic stress [52]; thus, increased GFAP expression indicates that cellular metabolic stress occurs in the brains of 6-month-old H67D mice.
However, at 12 months there is no longer increased cellular metabolic stress and indicates that the adaptive mechanisms to protect against iron-induced toxicity were successful.
However, the metabolic disruptions as consequences of the altered iron status were not completely restored given the finding of persistent alterations in cholesterol metabolism and behaviors [33] and increased ER stress [17] in the H67D mice.
In summary, there are two salient findings in this study. First, we provide the direct in vivo evidence that H63D HFE impacts brain iron homeostasis and creates environment for oxidative stress. Our findings together with recent MRI studies [10-13] shift the existing classical paradigm that brain is protected from iron overload associated with HFE mutations. Second, we demonstrated that as they age, H67D mice appear to develop adaptive mechanisms such as elevated Nrf2, decreased iron mobilization via Tf, and increased iron storage within L-ferritin that limits the amount of iron available to induce oxidative stress. Although there are compensatory mechanisms at younger ages, they are not sufficient to protect the brain because 6-month-old H67D mice have increased oxidative stress and astrogliosis. Moreover, our findings suggest that H67D mice may be more susceptible to environmental challenges and/or genetic modifiers at younger ages when there is a stress milieu and gliotic responses trying to adapt to the metabolic stress. The presence of oxidative and cellular stress in brains of H67D mice support our hypothesis that H63D HFE establishes an environment for pathogenic factors 115
that promote cellular damage and neurodegeneration. These data warrant the next line of investigation that I will report in the next chapter where a double mutant mouse line that carries both SOD1(G93A) mutation and H67D HFE has accelerated disease progression and shorter survival than SOD1(G93A) ALS mouse model [53]. Thus, these data strongly indicate that the H67D mouse line presented here can be used as a model to evaluate how
H63D HFE contributes to disease mechanisms and its impacts on the treatment strategies in neurodegenerative diseases.
2.6. Acknowledgements
This work is supported by the Judith and Jean Pape Adams Charitable Foundation and the George M. Leader Laboratory for Alzheimer’s disease research.
This chapter has previously been published and is reprinted from Biochimica et
Biophysica Acta – Molecular Basis of Disease, 1832, Nandar W, Neely EB, Unger EL,
Connor JR, A mutation in the HFE gene is associated with altered brain iron profiles and increased oxidative stress in mice, 729-741, Copyright © 2013, with permission from Elsevier B.V.
116
2.7. References
[1] P.A. Gochee, L.W. Powell, D.J. Cullen, D. Du Sart, E. Rossi, J.K. Olynyk, A
population-based study of the biochemical and clinical expression of the H63D
hemochromatosis mutation, Gastroenterology, 122 (2002) 646-651.
[2] M.E. Cogswell, M.L. Gallagher, K.K. Steinberg, D.S. Caudill Ph, A.C. Looker, B.A.
Bowman, E.W. Gunter, A.L. Franks, G.A. Satten, M.J. Khoury, L.M. Grummer-
Strawn, HFE genotype and transferrin saturation in the United States, Genet Med, 5
(2003) 304-310.
[3] O.T. Njajou, J.J. Houwing-Duistermaat, R.H. Osborne, N. Vaessen, J. Vergeer, J.
Heeringa, H.A. Pols, A. Hofman, C.M. van Duijn, A population-based study of the
effect of the HFE C282Y and H63D mutations on iron metabolism, Eur J Hum
Genet, 11 (2003) 225-231.
[4] W. Nandar, J.R. Connor, HFE gene variants affect iron in the brain, J Nutr, 141
(2011) 729S-739S.
[5] L. Batista-Nascimento, C. Pimentel, R.A. Menezes, C. Rodrigues-Pousada, Iron and
neurodegeneration: from cellular homeostasis to disease, Oxid Med Cell Longev,
2012 (2012) 128647.
[6] L. Zecca, M.B. Youdim, P. Riederer, J.R. Connor, R.R. Crichton, Iron, brain ageing
and neurodegenerative disorders, Nat Rev Neurosci, 5 (2004) 863-873.
[7] J.R. Connor, E.A. Milward, S. Moalem, M. Sampietro, P. Boyer, M.E. Percy, C.
Vergani, R.J. Scott, M. Chorney, Is hemochromatosis a risk factor for Alzheimer's
disease?, J Alzheimers Dis, 3 (2001) 471-477. 117
[8] J.H. Sheldon, Hemochromatosis, in, Oxford University Press, London, 1935, pp. 155-
159.
[9] J. Cammermyer, Deposition of iron in paraventricular areas of the human brain in
hemochromatosis, Journal of neuropathology and experimental neurology, 6 (1947)
111-127.
[10] J.E. Nielsen, L.N. Jensen, K. Krabbe, Hereditary haemochromatosis: a case of iron
accumulation in the basal ganglia associated with a parkinsonian syndrome, J Neurol
Neurosurg Psychiatry, 59 (1995) 318-321.
[11] D. Berg, U. Hoggenmuller, E. Hofmann, R. Fischer, M. Kraus, M. Scheurlen, G.
Becker, The basal ganglia in haemochromatosis, Neuroradiology, 42 (2000) 9-13.
[12] M.P. Rutgers, A. Pielen, M. Gille, Chronic cerebellar ataxia and hereditary
hemochromatosis: causal or coincidental association?, Journal of neurology, 254
(2007) 1296-1297.
[13] G. Bartzokis, P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P.
Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely,
J.R. Connor, Prevalent iron metabolism gene variants associated with increased
brain ferritin iron in healthy older men, J Alzheimers Dis, 20 (2010) 333-341.
[14] G. Bartzokis, P.H. Lu, K. Tingus, D.G. Peters, C.P. Amar, T.A. Tishler, J.P. Finn, P.
Villablanca, L.L. Altshuler, J. Mintz, E. Neely, J.R. Connor, Gender and iron genes
may modify associations between brain iron and memory in healthy aging,
Neuropsychopharmacology, 36 (2011) 1375-1384.
[15] N. Jahanshad, O. Kohannim, D.P. Hibar, J.L. Stein, K.L. McMahon, G.I. de
Zubicaray, S.E. Medland, G.W. Montgomery, J.B. Whitfield, N.G. Martin, M.J. 118
Wright, A.W. Toga, P.M. Thompson, Brain structure in healthy adults is related to
serum transferrin and the H63D polymorphism in the HFE gene, Proceedings of the
National Academy of Sciences of the United States of America, 109 (2012) E851-
859.
[16] S.Y. Lee, S.M. Patton, R.J. Henderson, J.R. Connor, Consequences of expressing
mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking
endogenous HFE, FASEB J, 21 (2007) 564-576.
[17] Y. Liu, S.Y. Lee, E. Neely, W. Nandar, M. Moyo, Z. Simmons, J.R. Connor, Mutant
HFE H63D protein is associated with prolonged endoplasmic reticulum stress and
increased neuronal vulnerability, The Journal of biological chemistry, 286 (2011)
13161-13170.
[18] R.M. Mitchell, S.Y. Lee, W.T. Randazzo, Z. Simmons, J.R. Connor, Influence of
HFE variants and cellular iron on monocyte chemoattractant protein-1, J
Neuroinflammation, 6 (2009) 6.
[19] R.M. Mitchell, S.Y. Lee, Z. Simmons, J.R. Connor, HFE polymorphisms affect
cellular glutamate regulation, Neurobiology of aging, 32 (2011) 1114-1123.
[20] E.C. Hall, 2nd, S.Y. Lee, N. Mairuae, Z. Simmons, J.R. Connor, Expression of the
HFE allelic variant H63D in SH-SY5Y cells affects tau phosphorylation at serine
residues, Neurobiology of aging, 32 (2011) 1409-1419.
[21] S. Tomatsu, K.O. Orii, R.E. Fleming, C.C. Holden, A. Waheed, R.S. Britton, M.A.
Gutierrez, S. Velez-Castrillon, B.R. Bacon, W.S. Sly, Contribution of the H63D
mutation in HFE to murine hereditary hemochromatosis, Proceedings of the National
Academy of Sciences of the United States of America, 100 (2003) 15788-15793. 119
[22] K.M. Erikson, D.J. Pinero, J.R. Connor, J.L. Beard, Regional brain iron, ferritin and
transferrin concentrations during iron deficiency and iron repletion in developing
rats, J Nutr, 127 (1997) 2030-2038.
[23] E.R. Stadtman, R.L. Levine, Protein oxidation, Ann N Y Acad Sci, 899 (2000) 191-
208.
[24] M. Conrad, H. Sato, The oxidative stress-inducible cystine/glutamate antiporter,
system x (c) (-) : cystine supplier and beyond, Amino Acids, 42 (2012) 231-246.
[25] A. Grochot-Przeczek, J. Dulak, A. Jozkowicz, Haem oxygenase-1: non-canonical
roles in physiology and pathology, Clin Sci (Lond), 122 (2012) 93-103.
[26] N. Freixenet, A. Remacha, E. Berlanga, A. Caixas, O. Gimenez-Palop, F. Blanco-
Vaca, V. Bach, M. Baiget, Y. Sanchez, J. Felez, J.M. Gonzalez-Clemente, Serum
soluble transferrin receptor concentrations are increased in central obesity. Results
from a screening programme for hereditary hemochromatosis in men with
hyperferritinemia, Clin Chim Acta, 400 (2009) 111-116.
[27] M. Jehn, J.M. Clark, E. Guallar, Serum ferritin and risk of the metabolic syndrome
in U.S. adults, Diabetes Care, 27 (2004) 2422-2428.
[28] E. Rossi, M.K. Bulsara, J.K. Olynyk, D.J. Cullen, L. Summerville, L.W. Powell,
Effect of hemochromatosis genotype and lifestyle factors on iron and red cell indices
in a community population, Clin Chem, 47 (2001) 202-208.
[29] S.H. Lee, J.W. Kim, S.H. Shin, K.P. Kang, H.C. Choi, S.H. Choi, K.U. Park, H.Y.
Kim, W. Kang, S.H. Jeong, HFE gene mutations, serum ferritin level, transferrin
saturation, and their clinical correlates in a Korean population, Dig Dis Sci, 54
(2009) 879-886. 120
[30] X.Y. Zhou, S. Tomatsu, R.E. Fleming, S. Parkkila, A. Waheed, J. Jiang, Y. Fei,
E.M. Brunt, D.A. Ruddy, C.E. Prass, R.C. Schatzman, R. O'Neill, R.S. Britton, B.R.
Bacon, W.S. Sly, HFE gene knockout produces mouse model of hereditary
hemochromatosis, Proceedings of the National Academy of Sciences of the United
States of America, 95 (1998) 2492-2497.
[31] M.S. Golub, S.L. Germann, R.S. Araiza, J.R. Reader, S.M. Griffey, K.C. Lloyd,
Movement disorders in the Hfe knockout mouse, Nutr Neurosci, 8 (2005) 239-244.
[32] D. Johnstone, R.M. Graham, D. Trinder, R.D. Delima, C. Riveros, J.K. Olynyk, R.J.
Scott, P. Moscato, E.A. Milward, Brain transcriptome perturbations in the Hfe(-/-)
mouse model of genetic iron loading, Brain Res, 1448 (2012) 144-152.
[33] F. Ali-Rahmani, P. Grigson, S. Lee, E. Neely, B. Kinsman, J.R. Connor, C.-L.
Schengrund, Effect of H63D-HFE on cholesterol metabolism, brain atrophy, and
cognitive impairment: implications for Alzheimer's Disease, (Submitted).
[34] S. Levi, A. Luzzago, G. Cesareni, A. Cozzi, F. Franceschinelli, A. Albertini, P.
Arosio, Mechanism of ferritin iron uptake: activity of the H-chain and deletion
mapping of the ferro-oxidase site. A study of iron uptake and ferro-oxidase activity
of human liver, recombinant H-chain ferritins, and of two H-chain deletion mutants,
The Journal of biological chemistry, 263 (1988) 18086-18092.
[35] P. Rucker, F.M. Torti, S.V. Torti, Role of H and L subunits in mouse ferritin, The
Journal of biological chemistry, 271 (1996) 33352-33357.
[36] T. Moos, Immunohistochemical localization of intraneuronal transferrin receptor
immunoreactivity in the adult mouse central nervous system, The Journal of
comparative neurology, 375 (1996) 675-692. 121
[37] T.K. Dickinson, J.R. Connor, Immunohistochemical analysis of transferrin receptor:
regional and cellular distribution in the hypotransferrinemic (hpx) mouse brain,
Brain Res, 801 (1998) 171-181.
[38] A.S. Dornelles, V.A. Garcia, M.N. de Lima, G. Vedana, L.A. Alcalde, M.R. Bogo,
N. Schroder, mRNA expression of proteins involved in iron homeostasis in brain
regions is altered by age and by iron overloading in the neonatal period, Neurochem
Res, 35 (2010) 564-571.
[39] B. Hallgren, P. Sourander, The effect of age on the non-haemin iron in the human
brain, Journal of neurochemistry, 3 (1958) 41-51.
[40] J.M. Hill, R.C. Switzer, 3rd, The regional distribution and cellular localization of
iron in the rat brain, Neuroscience, 11 (1984) 595-603.
[41] T. Moos, P.S. Oates, E.H. Morgan, Expression of the neuronal transferrin receptor is
age dependent and susceptible to iron deficiency, The Journal of comparative
neurology, 398 (1998) 420-430.
[42] D.F. Leitner, J.R. Connor, Functional roles of transferrin in the brain, Biochimica et
biophysica acta, 1820 (2012) 393-402.
[43] E.M. Taylor, A. Crowe, E.H. Morgan, Transferrin and iron uptake by the brain:
effects of altered iron status, Journal of neurochemistry, 57 (1991) 1584-1592.
[44] J.R. Burdo, D.A. Antonetti, E.B. Wolpert, J.R. Connor, Mechanisms and regulation
of transferrin and iron transport in a model blood-brain barrier system,
Neuroscience, 121 (2003) 883-890. 122
[45] J.R. Connor, B.S. Snyder, J.L. Beard, R.E. Fine, E.J. Mufson, Regional distribution
of iron and iron-regulatory proteins in the brain in aging and Alzheimer's disease, J
Neurosci Res, 31 (1992) 327-335.
[46] B. Todorich, X. Zhang, B. Slagle-Webb, W.E. Seaman, J.R. Connor, Tim-2 is the
receptor for H-ferritin on oligodendrocytes, Journal of neurochemistry, 107 (2008)
1495-1505.
[47] J.R. Burdo, J. Martin, S.L. Menzies, K.G. Dolan, M.A. Romano, R.J. Fletcher, M.D.
Garrick, L.M. Garrick, J.R. Connor, Cellular distribution of iron in the brain of the
Belgrade rat, Neuroscience, 93 (1999) 1189-1196.
[48] J.R. Burdo, S.L. Menzies, I.A. Simpson, L.M. Garrick, M.D. Garrick, K.G. Dolan,
D.J. Haile, J.L. Beard, J.R. Connor, Distribution of divalent metal transporter 1 and
metal transport protein 1 in the normal and Belgrade rat, J Neurosci Res, 66 (2001)
1198-1207.
[49] S. Levi, P. Santambrogio, A. Cozzi, E. Rovida, B. Corsi, E. Tamborini, S. Spada, A.
Albertini, P. Arosio, The role of the L-chain in ferritin iron incorporation. Studies of
homo and heteropolymers, J Mol Biol, 238 (1994) 649-654.
[50] J.R. Connor, K.L. Boeshore, S.A. Benkovic, S.L. Menzies, Isoforms of ferritin have
a specific cellular distribution in the brain, J Neurosci Res, 37 (1994) 461-465.
[51] E.A. Malecki, E.E. Cable, H.C. Isom, J.R. Connor, The lipophilic iron compound
TMH-ferrocene [(3,5,5-trimethylhexanoyl)ferrocene] increases iron concentrations,
neuronal L-ferritin, and heme oxygenase in brains of BALB/c mice, Biol Trace Elem
Res, 86 (2002) 73-84. 123
[52] J. Middeldorp, E.M. Hol, GFAP in health and disease, Prog Neurobiol, 93 (2011)
421-443.
[53] W. Nandar, E. Neely, Z. Simmons, J.R. Connor, The 21st international symposium
on ALS/MND, (2010).
124
Chapter 3
H63D HFE Accelerates Disease Progression in Animal Models of
Amyotrophic Lateral Sclerosis
3.1. Abstract
Despite increasing interest in the association between H63D HFE and amyotrophic lateral sclersosis (ALS), the impact of H63D HFE on ALS pathogenesis remains unclear. In this study, we generated a double mutant mouse line (SOD1/H67D) carrying H67D HFE (homologue of human H63D) and SOD1(G93A) mutations. We found the SOD1/H67D mice have a shorter survival and an accelerated disease progression. We examined a number of parameters in the lumbar spinal cord of
SOD1/H67D mice at 90 days (presymptomatic), 110 days (symptomatic) and end-stage.
Transferrin receptor and ferritin expression, both indicators of iron status were altered in
SOD1/H67D and SOD1 mice starting at 90 days. Transferrin receptor was decreased and ferritin was increased suggesting increased iron, However, SOD1/H67D mice had higher
L-ferritin expression than SOD1 mice suggesting higher iron in SOD1/H67D mice.
Increased L-ferritin and caspase-3 levels in the double mutant mice are also suggestive of microglial activation. Increased hemeoxygeanse-1 and decreased nuclear factor E2- related factor 2 (Nrf2) levels in SOD1/H67D compared to SOD1 mice strongly suggest the accelerated disease process could be associated with increased oxidative stress.
Although both SOD1 and SOD1/H67D mice had increased GFAP expression, the magnitude of an increase was higher in SOD1/H67D mice at 110 days suggesting gliosis was more active in these mice. Our findings support the argument that H63D HFE is a 125
disease modifier for ALS and suggest the pathway involved increased oxidative stress and gliosis.
3.2. Introduction
Amyotrophic lateral sclerosis (ALS), commonly known as Lou Gehrig’s disease, is characterized by degeneration of lower and upper motor neurons in the brainstem, spinal cord, and the motor cortex. The worldwide incidence of ALS is 1-2 per 100,000 and the average age of clinical onset is 55-60 years with an average survival of 3 to 5 year after symptom onset [1, 2]. However, the range of survival is from a few months to more than a decade after onset [3]. Because of the high variability in age of onset and in survival, ALS is proposed to be a heterogeneous disease. The majority of ALS cases (90-
95%) are sporadic (SALS), whereas 5-10% are inherited (familial ALS, or FALS).
Despite identification of mutations in a number of genes associated with FALS and
SALS [1, 2] including superoxide dismutase (SOD; [4]), TAR-DNA-binding protein 43
(TDP 43; [5]), fused in sarcoma/translated in liposarcoma (FUS/TLS; [6]) and chromosome 9 open reading frame 72 (C9ORF72; [7]), the etiology in most patients with
ALS remains inconclusive, and the molecular mechanisms contributing to motor neuron degeneration in ALS have not been elucidated.
Loss of iron homeostasis and the associated oxidative stress are significant parts of the disease processes in neurodegenerative diseases including ALS [8]. Higher iron levels in the central nervous system [9, 10] and elevated serum ferritin have been reported in ALS patients [11-13]. Treatment with iron chelators delayed onset, extended survival and prevented motor neuron degeneration in ALS mouse models [14, 15]. These 126
reports suggest an important role of iron metabolism in ALS pathogenesis. Therefore, we began studies to determine if there were polymorphisms associated with iron metabolism that could influence the ALS phenotype.
One of the genes involved in iron homeostasis is the HFE gene. Two common
HFE polymorphisms are H63D and C282Ywith worldwide allelic frequencies of 8.1% and 1.9% respectively. The C282Y HFE polymorphism is mostly associated with hereditary hemochromatosis (HH), the most common iron overload genetic disorder in
Caucasian population (1/200). The occurrence of the H63D HFE in HH is lower than
C282Y [16]. However, increasing evidence suggests an association of H63D HFE with neurodegenerative diseases [17] including ALS. Five independent groups in the United
States [18], the United Kingdom [19], Italy [20], the Netherlands [21] and China [22] have reported a positive association between H63D HFE variant and ALS. Although three studies [23-25] reported no association between H63D HFE and ALS, in all studies, there is agreement that H63D HFE is present in as many as 30% of ALS patients [18-21,
23-25]. Moreover, a meta-analysis indicates that the presence of the H63D HFE variant increases the risk of developing ALS by 4-fold [26].
The existing paradigm regarding HFE gene variants and brain function holds that the brain is protected from iron accumulation associated with the HFE polymorphisms because of the blood-brain-barrier. Recent MRI studies, however, suggest that people with HFE polymorphisms have more brain iron and increased cognitive impairment with age [27-30]. In an animal model, the presence of H67D HFE (homologous to H63D in human) disrupts brain iron homeostasis and is associated with increased oxidative stress in the brain [31]. The alterations in iron homeostasis and increased oxidative stress are 127
also seen at the cellular level [32], along with increased glutamate release [33] and increased endoplasmic reticulum (ER) stress [34]. Each of the above mechanisms is considered a contributing factor to ALS pathogenesis [1, 35]. Thus, the data strongly argue that H63D HFE is a genetic modifier for the risk of ALS, and warrant the development of an animal model as presented herein.
Based on findings from our previous in vitro and in vivo studies, we hypothesized that H63D HFE increases the risk of ALS by establishing a permissive milieu that promotes the convergence of disease mechanisms in ALS. To directly test our hypothesis, we generated a double mutant mouse line (SOD1/H67D) that carries both H67D HFE
(homologous to H63D in humans) and SOD1(G93A) mutations. We found that H67D
HFE shortens survival and disease duration in double mutant mice. Elevated oxidative stress, microglial toxicity and dysregulation of iron homeostasis contribute to an accelerated disease in these mice. Given the data that indicate 1/3 of patients with ALS carry the H63D gene variant, the double mutant mouse model could serve as a critical preclinical model to evaluate how the H63D HFE genotype can impact the disease process and treatment strategies for ALS patients.
3.3. Materials and Methods
3.3.1. Animal model: Double Mutant SOD1/H67D Mice Generation
SOD1(G93A) male mice (strain name: B6SJL-Tg(SOD1-G93A)1Gur/J; #002726) purchased from Jackson Labs (Bar Harbor, ME) were crossbred with H67D/H67D
(homologous to H63D in humans) or wild-type HFE female mice to generate a double mutant mouse line, that carries both H67D HFE and SOD1(G93A). SOD1(G93A) and 128
wild-type (WT) mice from the same litters as the double mutant mice were included in all experiments. Both males and females were included in all experiments.
Animals were maintained under normal housing conditions with ad libitum access to food and water. All experiments were performed according to the NIH Guide for the
Care and Use of Laboratory Animals and were approved by the Pennsylvania State
University College of Medicine Institutional Animal Care and Use Committee.
3.3.2. HFE and SOD1 (G93A) Genotyping
The H67D HFE genotyping was performed as previously reported [31]. Briefly,
DNA was extracted from tail biopsies using DNeasy blood and tissue kit (QIAGEN, CA).
PCR was performed using following forward and reverse primers:
(5′AGGACTCACTCTCTGGCAGCAGGAGGTAACCA3′) and
(5′TTTCTTTTACAAAGCTATATCCCCAGGGT3′). Following PCR, DNA was digested with BspHI restriction enzyme for 2 hours at 37ºC to detect H67D point mutation. Genotyping for SOD1(G93A) mutation was performed using primers that specifically amplifying a 236-bp DNA fragment carrying G93A mutation. The forward and reverse primers are: 5’CATCAGCCCTAATCCATCTGA-3’ and 5’-
CGCGACTAACAATCAAAGTGA-3’. PCR conditions for SOD1(G93A) genotyping are 95ºC for 3 minutes, 95ºC for 30 seconds, 60ºC for 30 seconds, 35 cycles of 72ºC for
45 seconds and 72ºC for 2 minutes. The PCR product was separated by 1.5% agarose gel electrophoresis.
129
3.3.3. Behavior and Survival
3.3.3.1. Rotarod
Starting at 49 days of age, motor performance was tested on a rotarod apparatus
(Columbus Instruments, Columbus, OH) rotating at 15 rpm. The amount of time that the mouse could stay on the rotarod before the first fall was recorded to determine disease onset. The duration of the rotarod test was 180 seconds, and was performed twice every week. A mouse was considered to fail the test when it could not stay on the rotarod for more than one standard error mean (> 1 SEM) below the mean time period it stayed on the rotarod during the presymptomatic phase. The probability of passing the rotarod test was analyzed by Kaplan-Meier (n = 19 to 32 per genotype).
3.3.3.2. Grip Strength
Hindlimb and forelimb strength were measured by a grip strength meter
(Columbus Instruments, Columbus, OH) to determine disease progression. Mice were held by the base of the tail and were allowed to grasp a horizontal metal bar attached to the grip strength meter with their forelimbs or hindlimbs. They were gently pulled back horizontally. Mice resist the increasing force by clinging onto the metal bar until they can no longer resist the force. The force applied to the bar at the moment the mouse released the bar was recorded as its maximum force. The test was repeated three times and an average determined for each animal. The grip test was performed once each week from
80 days to 127 days of age (n = 5 to 10 per genotype).
130
3.3.3.3. Survival and Disease Duration
End stage of the disease was defined as the inability of the animal to right itself within 30 seconds after being placed on its side. Kaplan-Meier survival analysis was performed to compare survival between experimental groups (n = 22 to 32 per genotype).
Disease duration was the mean time between age of disease onset, determined by rotarod test, and end stage of the disease (n = 19 to 32 per genotype).
3.3.4. Histology
Ninety- and 110-day-old mice (n = 6 to 9 per genotype) were perfused transcardially with Ringer’s solution followed by 4% paraformaldehyde in 0.1 M phosphate buffer. The spinal column was removed and post-fixed with 4% paraformaldehyde for 18 hours. The complete spinal cord was taken out and the lumbar region (L1-L5) was dissected. Paraffin embedded serial cross-sections (6 µm) of the lumbar spinal cords were processed for cresyl violet stain to determine motor neuron loss.
Briefly, the sections were deparaffinized and rehydrated through a series of ethanol, the sections were stained with 0.5% cresyl violet (in distilled water) for 7 minutes. Excess stain was rinsed in distilled water and 70% ethanol. Sections were dehydrated by immersing in 95% ethanol followed by 100% ethanol. Glacial acetic acid (in 95% ethanol) was included during the dehydration series to differentiate the stain. Motor neurons in anterior gray matter were counted every 10th section with bright-field microscopy by an investigator blinded for genotypes using the following criteria: 1) the presence of a large single nucleolus located within the nucleus and 2) a cell soma area over 100 µm2. We also counted larger motor neurons with a cell soma area over 250 µm2 131
[36]. A total of 10 sections were counted and averaged for each animal (n = 7 to 10 per genotype).
3.3.5. Western Blot
Lumbar spinal cord samples were harvested from presymptomatic age (90-day), symptomatic age (110-day) and end-stage SOD1/H67D mice (n = 6 to 10 per genotype).
Presymptomatic and symptomatic ages were chosen based on behavior studies. Samples from age-matched SOD1(G93A) and wild-type (WT) littermates were included in all of the analyses. Lumbar spinal cord tissues were homogenized in RIPA buffer (Sigma, St.
Louis, MO) with protease inhibitor cocktail (1:100; Sigma, St. Louis, MO). Total protein concentration was determined with Pierce BCA protein assay kit (Thermo scientific,
MA). Total spinal cord homogenates (20 µg total protein) was separated on 4-20%
Criterion polyacrylamide Tris-HCl gel (Bio-Rad, Hercules, CA) by electrophoresis and transferred overnight at 4ºC onto nitrocellulose membrane. For GAFP protein analysis, total protein of 10 µg was loaded. After blocking with 5% nonfat dry milk, the membranes were incubated with a primary antibody for overnight at 4ºC followed by an hour incubation with anti-host horseradish peroxidase-linked secondary antibodies
(Amersham Bioscience, Piscataway, NJ). The signal was visualized with enhanced chemiluminescent (ECL) system (Perkin Elmer, Waltham, MA) and the densitometric analysis was performed with Multigauge software (V3.0; Fuji film system).
Lumbar spinal cord samples were analyzed for expressions of H-ferritin (1:1000;
Covence, Princeton, NJ), L-ferritin (1:500; abcam, Cambridge, MA), transferrin receptor
(TfR, 1:500; Zymed Laboratories Inc., San Francisco, CA), hemeoxygenase-1 (HO-1, 132
1:500; Enzo Life Science, Farmingdale, NY), nuclear factor E2-related factor 2 (Nrf2,
1:1000; abcam, Cambridge, MA), total caspase-3 (1:500; Cell Signaling Technology Inc.,
Danvers, MA), GFAP (1:10,000; Dako, Carpinteria, CA) and beta-actin (1:3000; Sigma,
St. Louis, MO).
3.3.6. Statistical Analyses
Data were expressed as mean ± standard error. Kaplan-Meier was used to analyze the survival and probability of passing the rotarod test. An analysis of variance one-way
ANOVA or two-way ANOVA (GraphPad Prism 4; La Jolla, CA) followed by Tukey multiple comparison or Bonferroni posttest was used to compare between experimental groups. For Grip strength analysis, repeated measure mixed ANOVA with Tukey-Kramer posttest was performed using SAS 9.3 (Cary, NC). A p-value < 0.05 was considered significance for all the experiments.
3.4. Results
3.4.1 H67D HFE Shortens Survival and Accelerates Disease Progression
To determine the impact of H67D HFE on ALS disease pathogenesis, we generated a double mutant mouse line (SOD1/H67D) that carries H67D HFE and
SOD1(G93A) mutation (Figure 3-1).
133
Figure 3-1
Figure 3-1. A representative gel for genotyping of a double mutant mouse line
(SOD1/H67D). The H67D point mutation was detected by PCR amplification of genomic
DNA and subsequent digestion with BspHI restriction enzyme. DNA from WT mice digested by BspHI resulted 240 and 260 bp and DNA from +/H67D mice resulted 500,
240 and 260 bp (lane 1, 3, 5 and 7). Genotyping for SOD1(G93A) mutation was performed with primers specifically amplifying a 236-bp DNA fragment carrying G93A mutation (lane 2, 4, 6 and 8). The 100-bp marker was run at the beginning of the gel. The
SOD1/H67D mice are heterozygous for H67D HFE and also carry G93A mutation (lane
5 and 6) while SOD1(G93A ) mice carry wild-type HFE and G93A mutation (lane 1, 2 and lane 3, 4). 134
We determined survival, disease onset and disease progression in double mutant
(SOD1/H67D) mice and compared with SOD1 mice (G93A). The double mutant mice had significantly shorter survival compared to SOD1 mice (Figure 3-2 A). The median survival in double mutant and SOD1 mutant mice was 128 days and 132.5 days respectively (p = 0.02). We found a gender effect on both survival and disease duration.
The median survival of female double mutant mice (129 days) was significantly shorter than female SOD1 mice (137 days; p = 0.002; Figure 3-2 B). The median survival of male double mutant mice (128 days) was not different from male SOD1 mice (122 days; p = 0.17; Figures 3-2 C).
135
Figure 3-2
136
Figure 3-2. H67D HFE shortens survival in double mutant (SOD1/H67D) mice.
(A). Kaplan-Meier survival analysis comparing SOD1/H67D mice with SOD1(G93A)
mice. Median survival of SOD1/H67D mice is shorter when compared with
SOD1(G93A) mice (128 days vs. 132.5 days, n = 22 to 32 per genotype, p = 0.02).
(B). Median survival of females SOD1/H67D mice is shorter compared to female
SOD1(G93A) mice (129 vs. 137 days, n = 13 to 15 per genotype, p = 0.002).
(C). Median survival of males SOD1/H67D and males SOD1(G93A) mice is not different
(128 vs. 122 days, n = 9 to 17 per genotype, p = 0.17).
137
Disease duration, the time between disease onset and end-stage, was also shorter in double mutant mice (21 ± 1.8 days; mean ± SE) compared to SOD1 mice (26 ± 2.9 days); however, this difference did not reach statistical significance (p = 0.11; Figure 3-3
A). When disease duration was analyzed by gender, female double mutant mice (18 ± 1.4 days) exhibited significantly shorter disease duration than SOD1 mice (27 ± 3.6 days; mean ± SE; p = 0.02; Figure 3-3 B) though disease duration of male double mutant mice
(23 ± 3.3 days) was not different from male SOD1 mice (24 ± 4.9 days; p = 0.78; Figures
3-3 C).
138
Figure 3-3
Figure 3-3. Shorter disease duration in double mutant (SOD1/H67D) mice.
(A). The SOD1/H67D mice tend to have shorter disease duration, mean time between
disease onset and end-stage, compared to SOD1(G93A) mice (20.6 ± 1.9 days vs.
25.9 ± 2.8 days, n = 19 to 32, p = 0.11).
(B). There is a significantly shorter disease duration in females SOD1/H67D mice
compared to females SOD1(G93A) mice (18.1 ± 1.4 days vs. 26.9 ± 3.6 days, n = 11
to 15 per genotype, * p < 0.05).
(C). Disease duration in males SOD1/H67D is not different from male SOD1(G93A)
mice (22.8 ± 3.3 days vs. 24.5 ± 4.9 days, n = 8 to 17 per genotype, p = 0.78). 139
Age of disease onset in double mutant mice was not different from SOD1 mice
(Figure 3-4). The double mutant and SOD1 mice failed the rotarod test starting at 107 ±
1.9 days and 106 ± 2.9 days respectively (mean ± SE). The wild-type (WT) mice did not fail the rotarod test during the observation period (data not shown). There was no difference between the SOD1 and double mutant mice in age of disease onset as determined by performance on the rotarod (Figure 3-4).
140
Figure 3-4
Figure 3-4. Age of disease onset is not altered in double mutant (SOD1/H67D) mice (107
± 1.9 days vs. 106 ± 2.8 days). Kaplan-Meier analysis is used to compare the probability of passing the rotarod test. The first failure is defined as the inability of a mouse to stay on the rotarod for more than one standard error mean (> 1 s.e.m) below the mean time period it stayed on the rotarod during the presymtomatic phase. This is designated as the age of disease onset between groups. n = 19 to 32 mice per genotype, p = 0.77.
141
Significant weakness in forelimbs and hindlimbs was observed in both double mutant and SOD1 mice when compared with wild-type (WT) mice (Figure 3-5 A-B).
Compared to SOD1 mice, double mutant mice performed significantly worse on both forelimb and hindlimb grip strength. Indeed, the double mutant mice exhibit poorer forelimb and hindlimb strength than the SOD1 group beginning at the age of disease onset (106 days) until the end of the test (127 days) suggesting an accelerated disease progression in double mutant mice. There was no gender effect on age of disease onset or progression as measured by roatarod and grip strength.
142
Figure 3-5
143
Figure 3-5. Accelerated disease progression in double mutant (SOD1/H67D) mice. Grip strength meter was used to assess the strength of forelimbs and hindlimbs. The
SOD1/H67D mice have significantly less forelimb (A) and hindlimb strength (B) compared to SOD1(G93A) and WT mice. The SOD1(G93A) mice have lower forelimb and hindlimb strength than WT mice but stronger than SOD1/H67D mice. (** p < 0.01,
**** p < 0.0001; n = 5 to 10 per genotype).
144
3.4.2. H67D HFE has no Effect on Motor Neurons Loss
Even before disease onset at 90 days of age, a significant loss of motor neurons was already found in lumbar spinal cord of both the double mutant and the SOD1 mice
(27% compared to 21%), but the 6% greater loss of motor neurons in the double mutant mice was not statistically significant (Figure 3-6 A). The motor neuron loss continued to the 110 day time point, at which both double mutant and SOD1 mice had lost about half of the motor neurons compared to WT mice. There was no difference between double mutant and SOD1 mice (Figure 3-6 B). Large motor neurons loss was also present before disease onset at 90 days in both groups (Figure 3-6 C). At 110 days, both double mutant mice and SOD1 mice lost almost ¾ of large motor neurons compared to WT mice. There was no difference between the double mutant and SOD1 mice (Figure 3-6 D). No gender difference was observed between groups.
145
Figure 3-6
Figure 3-6. Motor neuron loss in SOD1(G93A) and SOD1/H67D mice. Bars represent mean ± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001 n = 7 to 10 per genotype).
(A-B). A significant loss of motor neurons (>100 µm2) is observed in SOD1/H67D and
SOD1(G93A) mice at 90 days and 110 days of age. Motor neuron loss in
SOD1/H67D mice do not differ from SOD1(G93A) mice.
(C-D). SOD1/H67D and SOD1(G93A) mice had a significant loss of large motor neurons
(>250 µm2) loss at 90 days and 110 days compared to WT mice. Large motor
neuron loss is not different between SOD1/H67D and SOD1(G93A) mice. 146
3.4.3. Altered Iron Management Protein Expression in Double Mutant (SOD1/H67D)
Mice
Because H67D HFE alters brain iron homeostasis [31], we proposed that altered iron homeostasis may contribute to shorter survival and accelerated disease observed in double mutant mice. Therefore, we determined the expression of proteins involved in iron homeostasis: transferrin receptor (TfR), L-ferritin and H-ferritin in lumbar spinal cords of
90 days, 110 days and end-stage double mutant and SOD1mice.
Compared to wild-type (WT) mice, a significant decrease in TfR expression, a cellular iron uptake protein, was found in double mutant and SOD1 mice starting at 90 days, and remained lower in both groups at 110 days and end-stage (Figure 3-7 A-C).
147
Figure 3-7
Figure 3-7. Decreased transferrin receptor (TfR) expression in lumbar spinal cords of double mutant (SOD1/H67D) mice. Lumbar spinal cord homogenates from 90-day-old,
110-day-old and end-stage SOD1/H67D, SOD1(G93A) and wild-type (WT) mice were used to determine the expression of transferrin receptor (TfR). A representative Western
Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph.
The expression level is normalized to β-actin. Transferrin receptor (TfR) expression in the lumbar spinal cords of SOD1/H67D and SOD1(G93A) mice is decreased at all ages compared to WT mice but TfR expression in double mutant and SOD1 mice are not different. Bars represent mean ± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001; n
= 6 to 10 per genotype). 148
Examination of the expression of L-ferritin, an iron storage protein particularly enriched in microglia, revealed a significant increase in the double mutant mice compared to WT and the SOD1 mutant mice at both 90 and 110 days (Figure 3-8 A-B).
Compared to WT mice, L-ferritin expression was 49% higher in double mutant mice at
90 days and remained increased with time to a maximal increase at end-stage (140% and
252% higher at 110 days and end-stage respectively). Notably, L-ferritin expression in double mutant mice was 86% and 79% higher than SOD1 mice at 90 days and 110 days but was similar between the two groups at end stage (Figure 3-8 A-C). The similar levels of L-ferritin at end stage between the two groups resulted from a dramatic increase in L- ferritin in the SOD1 mice to those levels seen in double mutant mice. L-ferritin expression in SOD1 mice was significantly increased compared to WT only at end-stage.
H-ferritin expression in both double mutant and SOD1mice were not different from WT mice at any age (Figure 3-9 A-C).
149
Figure 3-8
Figure 3-8. Increased L-ferritin expression in lumbar spinal cords of double mutant
(SOD1/H67D) mice. Lumbar spinal cord homogenates from 90-day-old, 110-day-old and end-stage SOD1/H67D, SOD1(G93A) and wild-type (WT) mice were used to determine the expression of L-ferritin. A representative Western Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph. The expression level is normalized to β-actin. L-ferritin expression is increased in the lumbar spinal cords of SOD1/H67D compared to SOD1(G93A) and wild-type mice at 90-days (A) and 110-days (B). At end- stage (C), SOD1/H67D as well as SOD1(G93A) mice have significantly higher L-ferritin expression than WT mice. L-ferritin expression in double mutant and SOD1 mice is not different at end-stage. Bars represent mean ± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001; n = 6 to 10 per genotype). 150
Figure 3-9
Figure 3-9. H-ferritin expression is not changed in lumbar spinal cords of double mutant
(SOD1/H67D) mice. Lumbar spinal cord homogenates from 90-day-old, 110-day-old and end-stage SOD1/H67D, SOD1(G93A) and wild-type (WT) mice were used to determine the expression of H-ferritin. A representative Western Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph. The expression level is normalized to β-actin. H-ferritin expression in SOD1/H67D and SOD1(G93A) mice does not differ from WT mice at all ages. Bars represent mean ± standard error. (n = 6 to 10 per genotype). 151
3.4.4. Increased Oxidative Stress in Double Mutant (SOD1/H67D) Mice
We determined expressions of hemeoxygenase-1 (HO-1) and nuclear factor E2- related factor 2 (Nrf2), as markers for oxidative stress in double mutant mice. HO-1 expression was significantly increased in double mutant mice compared to WT mice starting at 90 days and increased at each time point reaching maximal increase at end- stage (Figure 3-10 A-C). The expression of HO-1 was greater in the double mutant mice at 90 and 110 days compared with SOD1 mice (Figure 3-10 A and B). The HO-1 expression in SOD1 mice was significantly increased compared to WT only at end-stage
(Figure 3-10 C) and was a result of a 5-fold increase in the levels of HO-1 in the end- stage of the SOD1 mice.
Next, we determined Nrf2 expression in these mice to test the impact of genotype on an antioxidant system. Expression of Nrf2 in double mutant and SOD1 mice was not different from WT mice at 90 days; however, at both the symptomatic (110 days) and end-stage time periods both double mutant and SOD1 mice had significantly decreased
Nrf2 expression compared to WT mice (Figure 3-10 D-F). The Nrf2 expression in the lumbar spinal cords of double mutant mice was 20% less than that in SOD1 mice at symptomatic age (110 days) although this difference did not reach statistical significance
(Figure 3-10 E).
152
Figure 3-10
153
Figure 3-10. Increased oxidative stress in lumbar spinal cords of double mutant
(SOD1/H67D) mice. Lumbar spinal cord homogenates from 90-day-old, 110-day-old and end-stage SOD1/H67D, SOD1(G93A) and wild-type (WT) mice were determined for the expressions of hemeoxygenase-1 (HO-1) and nuclear factor E2-related factor 2 (Nrf2), as markers of oxidative stress. A representative Western Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph. The expression level is normalized to β-actin. Bars represent mean ± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001; n = 6 to 10 per genotype).
(A-C). Increased hemeoxygenase-1 (HO-1) expression is found in lumbar spinal cords of
SOD1/H67D compared to WT mice at 90 days, 110 days and end-stage. HO-1
expression in SOD1(G93A) mice is increased only at end-stage when compared to
WT mice.
(D-F). The Nrf2 expression is not different between groups at 90 days. Nrf2 expression is
decreased in SOD1/H67D and SOD1(G93A) mice at 110 days and at end-stage. At
110 days, Nrf2 expression is decreased by 20% in SOD1/H67D compared to
SOD1(G93A) mice; whereas, Nrf2 expression in both groups is not different at
end-stage. 154
3.4.5. Increased Caspase-3 in Double Mutant (SOD1/H67D) Mice
Caspase-3 expression in the lumbar spinal cords was not different between the three groups at 90 days (Figure 3-11 A). However, at symptomatic age (110 days), caspase-3 expression was significantly increased in the double mutant mice compared to
WT and SOD1 mice (Figure 3-11 B). At end-stage, caspase-3 expression was significantly increased in both double mutant and SOD1 mice compared to WT mice.
Caspase-3 expression was not different between double mutant and SOD1 mice at end- stage (Figure 3-11 C).
155
Figure 3-11
Figure 3-11. Increased caspase-3 expression in lumbar spinal cords of double mutant
(SOD1/H67D) mice. At 90 days, caspase-3 expression in lumbar spinal cords was not different between groups. At 110 days, SOD1/H67D mice has a significant increase in caspase-3 expression compared to WT and SOD1(G93A) mice. At end-stage, caspase-3 expression in both SOD1/H67D and SOD1(G93A) mice is increased compared to WT mice but caspase-3 expression in SOD1/H67D and SOD1(G93A) mice is not different. A representative Western Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph. The expression level is normalized to β-actin. Bars represent mean
± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001; n = 6 to 10 per genotype).
156
3.4.6. Increased Gliosis in Double Mutant (SOD1/H67D) Mice
A significant increase in total GFAP expression was observed in lumbar spinal cords of both double mutant and SOD1 mice starting at 90 days, presymptomatic age and remained higher at 110 days and at end-stage in both groups (Figure 3-12 A-C). At symptomatic age (110 days), the magnitude of increased GFAP expression was higher in the double mutant mice than SOD1 mice (126% increase vs. 78% increase compared to
WT respectively; Figure 3-12 B) but the differences between the double mutant and the
SOD1 did not reach statistical significance.
157
Figure 3-12
Figure 3-12. Increased GFAP expression in lumbar spinal cords of double mutant
(SOD1/H67D) mice. A significant increase in GFAP expression is observed in lumbar spinal cords of both SOD1/H67D and SOD1(G93A) mice starting at 90 days and remains increased at 110 days and end-stage. At 110 days, a magnitude of an increased in GFAP expression in double mutant mice is higher compared with SOD1(G93A) mice. A representative Western Blot gel is shown for each protein and the quantitative of blots is shown as a bar graph. The expression level is normalized to β-actin. Bars represent mean
± standard error. (* p < 0.05, ** p < 0.01, *** p < 0.001; n = 6 to 10 per genotype).
158
Table 3-1
Protein expression relative to the wild-type mice
90-days (presymptomatic Proteins 110-days (symptomatic age) End-stage age) SOD1(G93A) SOD1/H67D SOD1(G93A) SOD1/H67D SOD1(G93A) SOD1/H67D
Iron management proteins
H-ferritin NS NS NS NS NS NS
L-ferritin NS NS
Transferrin receptor (TfR)
Oxidative stress
Nuclear factor E2-related factor NS NS 2 (Nrf2) Hemeoxygenase- NS NS 1 (HO-1)
Gliosis
Caspase-3 NS NS NS (microgliosis)
L-ferritin NS NS (microgliosis)
GFAP (astrogliosis)
Table 3-1. Summary of findings demonstrating pathogenic pathways underlying accelerated disease in double mutant mice 159
3.5. Discussion
In this study, we demonstrated that H63D HFE genotype shortens survival and disease duration in double mutant mice. The presence of the H67D polymorphism altered iron homeostatic mechanism in the spinal cord of the mice as demonstrated by decreased
TfR and increased L-ferritin expression. The alterations in iron homeostasis in the double mutant model may contribute to the increased oxidative stress, which in turn would support the gliosis and increased expression of caspase-3. Together we conclude that elevated oxidative stress, increased microglia activation and altered iron homeostasis are underlying mechanisms contributing to accelerated disease in double mutant
(SOD1/H67D) mice. These results support our hypothesis that H63D HFE is a genetic modifier of ALS and modifies the disease by creating an environment that promotes the convergence of disease processes in ALS.
There is increasing evidence suggesting an association between H63D HFE and
ALS [18-22]. Even those studies in which a significant increase in H63D HFE was not found in the ALS population compared to control [23-25] the percentage of ALS patients with H63D HFE is consistently reported at around 30%. Only one other gene variant [37] has been reported to have higher frequency than H63D HFE in the ALS but this finding has not been confirmed.
However, the impact of the HFE genotype on ALS pathogenesis is less established. We demonstrated that double mutant (SOD1/H67D) mice, which are heterozygous for H67D HFE and also carry the SOD1(G93A) mutation, have a shorter survival and an accelerated disease progression. Genetic background can influence disease onset, severity of disease and survival in ALS rodent models independent of 160
transgene copy numbers [38-40]. The double mutant mice are on a C57BL6 background and this genetic background is associated with longer survival and milder disease phenotype [38-40], which is opposite from what we observed in double mutant mice.
Therefore, our results strongly suggest that H63D HFE is a contributing factor in ALS disease pathogenesis. In addition, we found gender influences on survival in an ALS animal model. We found female SOD1 mice survive longer than males, which is consistent with previous studies [38, 41]. However, after ovariectomy, survival in females and males SOD1 mice were not different [41] suggesting that increased lifespan in females SOD1 mice may be due to protective effects of estrogen.
The double mutant model was examined to provide insights into mechanisms by which H63D HFE contributes to accelerated disease progression in ALS. Transferrin receptor (TfR) expression was decreased in lumbar spinal cord of double mutant mice starting from presymptomatic age and remained lower until end-stage. Because TfR is expressed predominantly on neurons [42-44] the decrease could reflect the loss of neurons in the SOD1 and double mutant mice. However, TfR is a cellular iron uptake protein and its expression is post-transcriptionally regulated by cellular iron status.
Excess iron downregulates TfR protein synthesis while iron deficiency increases TfR synthesis [8, 45]. Thus, decreased TfR expression in double mutant mice could also reflect increase iron load in the neurons of the double mutant mice.
Similar to TfR, H-ferritin, an iron storage protein, is mainly expressed in neurons
[46]. Even though the double mutant mice have lost about half of motor neurons compared to WT mice, total H-ferritin levels are not changed significantly. This suggests
H-ferritin expression may be increased in surviving neurons, reflecting a neuronal iron 161
accumulation in the double mutant mice. Moreover, L-ferritin expression is also dramatically increased in the double mutant mice even before disease onset. L-ferritin is a long-term iron storage protein primarily enriched in microglia [46]. Ferritin synthesis is post-transcriptionally regulated by iron status; iron overload increases synthesis of ferritin
[45]. Therefore, higher L-ferritin levels in double mutant mice compared to SOD1 mice further suggests iron accumulation in the spinal cord of these mice.
In addition to iron loading, pro-inflammatory cytokines associated with activated microglia can increase L-ferritin expression in these cells [47]. Because neuroinflammation is a pathological hallmark in ALS and increased inflammatory cytokines together with microgliosis are present in ALS [48, 49], the increase in L- ferritin suggests that additional influences, such as microglial toxicity or inflammation, may contribute to accelerated disease in double mutant mice. Extracellular ferritin has been identified in two studies [11, 13] as being elevated in ALS and having a negative correlation to survival. The relationship of extracellular ferritin to our findings is not known at present but it is tempting to speculate that the increase in extracellular ferritin reflects activation of microglia and their increased expression and release of L-ferritin.
Ferritin release by microglia has been established [50].
Alterations in iron homeostasis and inflammatory environments are associated with, oxidative stress which is a significant part of the disease processes in ALS. We have previously demonstrated that the presence of H63D HFE creates an environment for oxidative stress [31, 32]. In this study, oxidative stress, indicated by increased HO-1 and decreased Nrf2, is present in both double mutant and SOD1 mice. However, the magnitude of the changes particularly at the symptomatic stage, are greater in double 162
mutant mice compared to SOD1 mice. The decrease in Nrf2, which regulates cellular antioxidant responses [51] in double mutant mice compared to SOD1 mice is particularly noteworthy as potentially part of the pathogenesis of the disease. Previous studies have reported decreased Nrf2 expression in postmortem brain and spinal cord tissues from sporadic ALS patients [52] and in SOD1(G93A) motor neurons [53]. Decreased Nrf2 also increases motor neurons sensitivity to nerve grown factor induced apoptosis [53].
Upregulating antioxidant genes by Nrf2 activators slow disease progression in an ALS mouse model [54]. These reports together with our findings suggest that the decreased
Nrf2 levels in double mutant mice compared to the SOD1 mice may be a driver of the accelerated disease in the double mutant mice possibly by reducing glutathione (GSH) level, an important anti-oxidant protein downstream of the Nrf2 [51, 55]. Indeed, decreased GSH in an ALS mouse model accelerates disease and shortens survival due to increased oxidative stress; phenotypes similar to those that we observed in double mutant mice in this study.
Moreover, double mutant mice have increased caspase-3 expression without additional motor neurons loss compared to SOD1 mice. Although caspase-3 activation is well accepted as a downstream event in the apoptotic pathway, a new non-apoptotic role of caspase-3 in regulation microglial activation and neurotoxicity has been reported.
Activating microglia with immunogens such as lipopolysaccharide increases expressions of caspase-3,-7 and -8 without microglial cell death. By inhibiting caspases, microglial activation is blocked and its associated neuronal toxicity is reduced [56]. Therefore, increased caspase-3 in double mutant mice may be associated with increased microglial toxicity, which is consistent with increased L-ferritin levels in these mice. Given that 163
damage in microglia mainly affects disease progression; whereas motor neurons toxicity affects disease onset [57, 58], accelerated disease progression together with increased L- ferritin and caspase-3 in double mutant mice strongly implicate increased microglial toxicity in accelerated disease progression in the double transgenic mice.
An additional change in glia in double mutant mice is astrogliosis indicated by increased GFAP expression. Astrogilosis is a well-known pathological hallmark in ALS.
Similar to microglia, damage in astrocytes mainly affects disease progression without influencing disease onset in ALS. In ALS animal models astrogliosis appears before disease onset [58]. Consistently, in our study, increased GFAP expression is present in both double mutant and SOD1 mice before disease onset and is progressively increased; however, the magnitude of an increase in GFAP is higher in double mutant mice, particularly at the symptomatic stage. Activated astrocytes can mediate motor neuron toxicity by secreting uncharacterized soluble factors toxic to motor neurons [59] and by enhancing microglial inflammatory responses, including production of nitric oxide and cytokines [58]. Therefore, elevated astrogliosis in the double mutant mice compared to
SOD1 mice may contribute to increased neuronal toxicity and microglial inflammatory responses, further contributing to an accelerate disease in double mutant mice.
In summary, we determined the impact of H63D HFE on ALS pathogenesis by generating a double mutant mouse line (SOD1/H67D). There are two important findings in this study. First, H63D HFE shortens survival and accelerates disease progression in double mutant mice. Importantly, SOD1/H67D mice have accelerated disease progression in the absence of any environmental stressors or challenges that may impact
ALS onset such as inflammation, diet or exposure to toxins. Secondly, our findings 164
indicate that significant disruption in cellular functions associated with H63D HFE is more extensive than the SOD(G93A) alone. The addition of the HFE H63D gene variant to the SOD1 mouse model of ALS increased oxidative stress and microglial toxicity, which are underlying mechanisms causing shorter survival and accelerated disease progression. Because H63D HFE is present in as many as 30% of ALS patients, findings from SOD1/H67D mice could have meaningful clinical implications such as a preclinical model in which to test how H63D HFE can impact treatment strategies in neurodegenerative diseases. Lastly, this study is one example of using SOD1/H67D mice as a model to determine which pathways that, when activated or deactivated lead to accelerated disease progression in ALS. 165
3.6. Acknowledgements
This work is supported by Judith and Jean Pape Adams Charitable Foundation, the
Paul and Harriett Campbell Fund for ALS research, Zimmerman Family Love Fund and the Robert Luongo ALS Fund.
166
3.7. References
[1] L. Ferraiuolo, J. Kirby, A.J. Grierson, M. Sendtner, P.J. Shaw, Molecular pathways of
motor neuron injury in amyotrophic lateral sclerosis, Nat Rev Neurol, 7 (2011) 616-
630.
[2] W. Robberecht, T. Philips, The changing scene of amyotrophic lateral sclerosis, Nat
Rev Neurosci, 14 (2013) 248-264.
[3] L. Regal, L. Vanopdenbosch, P. Tilkin, L. Van den Bosch, V. Thijs, R. Sciot, W.
Robberecht, The G93C mutation in superoxide dismutase 1: clinicopathologic
phenotype and prognosis, Archives of neurology, 63 (2006) 262-267.
[4] D.R. Rosen, T. Siddique, D. Patterson, D.A. Figlewicz, P. Sapp, A. Hentati, D.
Donaldson, J. Goto, J.P. O'Regan, H.X. Deng, et al., Mutations in Cu/Zn superoxide
dismutase gene are associated with familial amyotrophic lateral sclerosis, Nature,
362 (1993) 59-62.
[5] M. Neumann, D.M. Sampathu, L.K. Kwong, A.C. Truax, M.C. Micsenyi, T.T. Chou,
J. Bruce, T. Schuck, M. Grossman, C.M. Clark, L.F. McCluskey, B.L. Miller, E.
Masliah, I.R. Mackenzie, H. Feldman, W. Feiden, H.A. Kretzschmar, J.Q.
Trojanowski, V.M. Lee, Ubiquitinated TDP-43 in frontotemporal lobar degeneration
and amyotrophic lateral sclerosis, Science, 314 (2006) 130-133.
[6] T.J. Kwiatkowski, Jr., D.A. Bosco, A.L. Leclerc, E. Tamrazian, C.R. Vanderburg, C.
Russ, A. Davis, J. Gilchrist, E.J. Kasarskis, T. Munsat, P. Valdmanis, G.A. Rouleau,
B.A. Hosler, P. Cortelli, P.J. de Jong, Y. Yoshinaga, J.L. Haines, M.A. Pericak-
Vance, J. Yan, N. Ticozzi, T. Siddique, D. McKenna-Yasek, P.C. Sapp, H.R.
Horvitz, J.E. Landers, R.H. Brown, Jr., Mutations in the FUS/TLS gene on 167
chromosome 16 cause familial amyotrophic lateral sclerosis, Science, 323 (2009)
1205-1208.
[7] M. DeJesus-Hernandez, I.R. Mackenzie, B.F. Boeve, A.L. Boxer, M. Baker, N.J.
Rutherford, A.M. Nicholson, N.A. Finch, H. Flynn, J. Adamson, N. Kouri, A.
Wojtas, P. Sengdy, G.Y. Hsiung, A. Karydas, W.W. Seeley, K.A. Josephs, G.
Coppola, D.H. Geschwind, Z.K. Wszolek, H. Feldman, D.S. Knopman, R.C.
Petersen, B.L. Miller, D.W. Dickson, K.B. Boylan, N.R. Graff-Radford, R.
Rademakers, Expanded GGGGCC hexanucleotide repeat in noncoding region of
C9ORF72 causes chromosome 9p-linked FTD and ALS, Neuron, 72 (2011) 245-
256.
[8] S. Oshiro, M.S. Morioka, M. Kikuchi, Dysregulation of iron metabolism in
Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis, Adv
Pharmacol Sci, 2011 (2011) 378278.
[9] H. Oba, T. Araki, K. Ohtomo, S. Monzawa, G. Uchiyama, K. Koizumi, Y. Nogata, K.
Kachi, Z. Shiozawa, M. Kobayashi, Amyotrophic lateral sclerosis: T2 shortening in
motor cortex at MR imaging, Radiology, 189 (1993) 843-846.
[10] E.J. Kasarskis, L. Tandon, M.A. Lovell, W.D. Ehmann, Aluminum, calcium, and
iron in the spinal cord of patients with sporadic amyotrophic lateral sclerosis using
laser microprobe mass spectroscopy: a preliminary study, Journal of the neurological
sciences, 130 (1995) 203-208.
[11] E.F. Goodall, M.S. Haque, K.E. Morrison, Increased serum ferritin levels in
amyotrophic lateral sclerosis (ALS) patients, Journal of neurology, 255 (2008) 1652-
1656. 168
[12] M. Qureshi, R.H. Brown, Jr., J.T. Rogers, M.E. Cudkowicz, Serum ferritin and metal
levels as risk factors for amyotrophic lateral sclerosis, Open Neurol J, 2 (2008) 51-
54.
[13] Y. Nadjar, P. Gordon, P. Corcia, G. Bensimon, L. Pieroni, V. Meininger, F.
Salachas, Elevated serum ferritin is associated with reduced survival in amyotrophic
lateral sclerosis, PloS one, 7 (2012) e45034.
[14] S.Y. Jeong, K.I. Rathore, K. Schulz, P. Ponka, P. Arosio, S. David, Dysregulation of
iron homeostasis in the CNS contributes to disease progression in a mouse model of
amyotrophic lateral sclerosis, The Journal of neuroscience : the official journal of the
Society for Neuroscience, 29 (2009) 610-619.
[15] Q. Wang, X. Zhang, S. Chen, S. Zhang, M. Youdium, W. Le, Prevention of motor
neuron degeneration by novel iron chelators in SOD1(G93A) transgenic mice of
amyotrophic lateral sclerosis, Neurodegener Dis, 8 (2011) 310-321.
[16] A.T. Merryweather-Clarke, J.J. Pointon, J.D. Shearman, K.J. Robson, Global
prevalence of putative haemochromatosis mutations, J Med Genet, 34 (1997) 275-
278.
[17] W. Nandar, J.R. Connor, HFE gene variants affect iron in the brain, J Nutr, 141
(2011) 729S-739S.
[18] X.S. Wang, S. Lee, Z. Simmons, P. Boyer, K. Scott, W. Liu, J. Connor, Increased
incidence of the Hfe mutation in amyotrophic lateral sclerosis and related cellular
consequences, Journal of the neurological sciences, 227 (2004) 27-33. 169
[19] E.F. Goodall, M.J. Greenway, I. van Marion, C.B. Carroll, O. Hardiman, K.E.
Morrison, Association of the H63D polymorphism in the hemochromatosis gene
with sporadic ALS, Neurology, 65 (2005) 934-937.
[20] G. Restagno, F. Lombardo, P. Ghiglione, A. Calvo, E. Cocco, L. Sbaiz, R. Mutani,
A. Chio, HFE H63D polymorphism is increased in patients with amyotrophic lateral
sclerosis of Italian origin, Journal of neurology, neurosurgery, and psychiatry, 78
(2007) 327.
[21] N.A. Sutedja, R.J. Sinke, P.W. Van Vught, M.W. Van der Linden, J.H. Wokke,
C.M. Van Duijn, O.T. Njajou, Y.T. Van der Schouw, J.H. Veldink, L.H. Van den
Berg, The association between H63D mutations in HFE and amyotrophic lateral
sclerosis in a Dutch population, Archives of neurology, 64 (2007) 63-67.
[22] X. He, X. Lu, J. Hu, J. Xi, D. Zhou, H. Shang, L. Liu, H. Zhou, B. Yan, L. Yu, F.
Hu, Z. Liu, L. He, X. Yao, Y. Xu, H63D polymorphism in the hemochromatosis
gene is associated with sporadic amyotrophic lateral sclerosis in China, European
journal of neurology : the official journal of the European Federation of
Neurological Societies, 18 (2011) 359-361.
[23] A.A. Yen, E.P. Simpson, J.S. Henkel, D.R. Beers, S.H. Appel, HFE mutations are
not strongly associated with sporadic ALS, Neurology, 62 (2004) 1611-1612.
[24] J. Praline, H. Blasco, P. Vourc'h, V. Rat, C. Gendrot, W. Camu, C.R. Andres, Study
of the HFE gene common polymorphisms in French patients with sporadic
amyotrophic lateral sclerosis, Journal of the neurological sciences, 317 (2012) 58-61.
[25] W. van Rheenen, F.P. Diekstra, P.T. van Doormaal, M. Seelen, K. Kenna, R.
McLaughlin, A. Shatunov, D. Czell, M.A. van Es, P.W. van Vught, P. van Damme, 170
B.N. Smith, S. Waibel, H.J. Schelhaas, A.J. van der Kooi, M. de Visser, M. Weber,
W. Robberecht, O. Hardiman, P.J. Shaw, C.E. Shaw, K.E. Morrison, A. Al-Chalabi,
P.M. Andersen, A.C. Ludolph, J.H. Veldink, L.H. van den Berg, H63D
polymorphism in HFE is not associated with amyotrophic lateral sclerosis,
Neurobiology of aging, 34 (2013) 1517 e1515-1517.
[26] C. Ellervik, H. Birgens, A. Tybjaerg-Hansen, B.G. Nordestgaard, Hemochromatosis
genotypes and risk of 31 disease endpoints: meta-analyses including 66,000 cases
and 226,000 controls, Hepatology, 46 (2007) 1071-1080.
[27] J.E. Nielsen, L.N. Jensen, K. Krabbe, Hereditary haemochromatosis: a case of iron
accumulation in the basal ganglia associated with a parkinsonian syndrome, Journal
of neurology, neurosurgery, and psychiatry, 59 (1995) 318-321.
[28] D. Berg, U. Hoggenmuller, E. Hofmann, R. Fischer, M. Kraus, M. Scheurlen, G.
Becker, The basal ganglia in haemochromatosis, Neuroradiology, 42 (2000) 9-13.
[29] G. Bartzokis, P.H. Lu, T.A. Tishler, D.G. Peters, A. Kosenko, K.A. Barrall, J.P.
Finn, P. Villablanca, G. Laub, L.L. Altshuler, D.H. Geschwind, J. Mintz, E. Neely,
J.R. Connor, Prevalent iron metabolism gene variants associated with increased
brain ferritin iron in healthy older men, J Alzheimers Dis, 20 (2010) 333-341.
[30] G. Bartzokis, P.H. Lu, K. Tingus, D.G. Peters, C.P. Amar, T.A. Tishler, J.P. Finn, P.
Villablanca, L.L. Altshuler, J. Mintz, E. Neely, J.R. Connor, Gender and iron genes
may modify associations between brain iron and memory in healthy aging,
Neuropsychopharmacology, 36 (2011) 1375-1384. 171
[31] W. Nandar, E.B. Neely, E. Unger, J.R. Connor, A mutation in the HFE gene is
associated with altered brain iron profiles and increased oxidative stress in mice,
Biochimica et biophysica acta, 1832 (2013) 729-741.
[32] S.Y. Lee, S.M. Patton, R.J. Henderson, J.R. Connor, Consequences of expressing
mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking
endogenous HFE, FASEB J, 21 (2007) 564-576.
[33] R.M. Mitchell, S.Y. Lee, Z. Simmons, J.R. Connor, HFE polymorphisms affect
cellular glutamate regulation, Neurobiology of aging, 32 (2011) 1114-1123.
[34] Y. Liu, S.Y. Lee, E. Neely, W. Nandar, M. Moyo, Z. Simmons, J.R. Connor, Mutant
HFE H63D protein is associated with prolonged endoplasmic reticulum stress and
increased neuronal vulnerability, The Journal of biological chemistry, 286 (2011)
13161-13170.
[35] J.D. Rothstein, Current hypotheses for the underlying biology of amyotrophic lateral
sclerosis, Annals of neurology, 65 Suppl 1 (2009) S3-9.
[36] D.B. Drachman, K. Frank, M. Dykes-Hoberg, P. Teismann, G. Almer, S.
Przedborski, J.D. Rothstein, Cyclooxygenase 2 inhibition protects motor neurons
and prolongs survival in a transgenic mouse model of ALS, Annals of neurology, 52
(2002) 771-778.
[37] L.S. Ro, S.L. Lai, C.M. Chen, S.T. Chen, Deleted 4977-bp mitochondrial DNA
mutation is associated with sporadic amyotrophic lateral sclerosis: a hospital-based
case-control study, Muscle & nerve, 28 (2003) 737-743.
[38] T.D. Heiman-Patterson, J.S. Deitch, E.P. Blankenhorn, K.L. Erwin, M.J. Perreault,
B.K. Alexander, N. Byers, I. Toman, G.M. Alexander, Background and gender 172
effects on survival in the TgN(SOD1-G93A)1Gur mouse model of ALS, Journal of
the neurological sciences, 236 (2005) 1-7.
[39] T.D. Heiman-Patterson, R.B. Sher, E.A. Blankenhorn, G. Alexander, J.S. Deitch,
C.B. Kunst, N. Maragakis, G. Cox, Effect of genetic background on phenotype
variability in transgenic mouse models of amyotrophic lateral sclerosis: a window of
opportunity in the search for genetic modifiers, Amyotrophic lateral sclerosis :
official publication of the World Federation of Neurology Research Group on Motor
Neuron Diseases, 12 (2011) 79-86.
[40] R. Mancuso, S. Olivan, P. Mancera, A. Pasten-Zamorano, R. Manzano, C. Casas, R.
Osta, X. Navarro, Effect of genetic background on onset and disease progression in
the SOD1-G93A model of amyotrophic lateral sclerosis, Amyotrophic lateral
sclerosis : official publication of the World Federation of Neurology Research Group
on Motor Neuron Diseases, 13 (2012) 302-310.
[41] C.I. Choi, Y.D. Lee, B.J. Gwag, S.I. Cho, S.S. Kim, H. Suh-Kim, Effects of estrogen
on lifespan and motor functions in female hSOD1 G93A transgenic mice, Journal of
the neurological sciences, 268 (2008) 40-47.
[42] T. Moos, Immunohistochemical localization of intraneuronal transferrin receptor
immunoreactivity in the adult mouse central nervous system, The Journal of
comparative neurology, 375 (1996) 675-692.
[43] T.K. Dickinson, J.R. Connor, Immunohistochemical analysis of transferrin receptor:
regional and cellular distribution in the hypotransferrinemic (hpx) mouse brain,
Brain Res, 801 (1998) 171-181. 173
[44] T. Moos, P.S. Oates, E.H. Morgan, Expression of the neuronal transferrin receptor is
age dependent and susceptible to iron deficiency, The Journal of comparative
neurology, 398 (1998) 420-430.
[45] L. Zecca, M.B. Youdim, P. Riederer, J.R. Connor, R.R. Crichton, Iron, brain ageing
and neurodegenerative disorders, Nat Rev Neurosci, 5 (2004) 863-873.
[46] J.R. Connor, K.L. Boeshore, S.A. Benkovic, S.L. Menzies, Isoforms of ferritin have
a specific cellular distribution in the brain, Journal of neuroscience research, 37
(1994) 461-465.
[47] K.I. Rathore, A. Redensek, S. David, Iron homeostasis in astrocytes and microglia is
differentially regulated by TNF-alpha and TGF-beta1, Glia, 60 (2012) 738-750.
[48] S.H. Appel, W. Zhao, D.R. Beers, J.S. Henkel, The microglial-motoneuron dialogue
in ALS, Acta Myol, 30 (2011) 4-8.
[49] R.M. Mitchell, Z. Simmons, J.L. Beard, H.E. Stephens, J.R. Connor, Plasma
biomarkers associated with ALS and their relationship to iron homeostasis, Muscle
& nerve, 42 (2010) 95-103.
[50] X. Zhang, N. Surguladze, B. Slagle-Webb, A. Cozzi, J.R. Connor, Cellular iron
status influences the functional relationship between microglia and
oligodendrocytes, Glia, 54 (2006) 795-804.
[51] S. Petri, S. Korner, M. Kiaei, Nrf2/ARE Signaling Pathway: Key Mediator in
Oxidative Stress and Potential Therapeutic Target in ALS, Neurol Res Int, 2012
(2012) 878030.
[52] A. Sarlette, K. Krampfl, C. Grothe, N. Neuhoff, R. Dengler, S. Petri, Nuclear
erythroid 2-related factor 2-antioxidative response element signaling pathway in 174
motor cortex and spinal cord in amyotrophic lateral sclerosis, Journal of
neuropathology and experimental neurology, 67 (2008) 1055-1062.
[53] M. Pehar, M.R. Vargas, K.M. Robinson, P. Cassina, P.J. Diaz-Amarilla, T.M.
Hagen, R. Radi, L. Barbeito, J.S. Beckman, Mitochondrial superoxide production
and nuclear factor erythroid 2-related factor 2 activation in p75 neurotrophin
receptor-induced motor neuron apoptosis, The Journal of neuroscience : the official
journal of the Society for Neuroscience, 27 (2007) 7777-7785.
[54] A. Neymotin, N.Y. Calingasan, E. Wille, N. Naseri, S. Petri, M. Damiano, K.T.
Liby, R. Risingsong, M. Sporn, M.F. Beal, M. Kiaei, Neuroprotective effect of
Nrf2/ARE activators, CDDO ethylamide and CDDO trifluoroethylamide, in a mouse
model of amyotrophic lateral sclerosis, Free Radic Biol Med, 51 (2011) 88-96.
[55] T.W. Kensler, N. Wakabayashi, S. Biswal, Cell survival responses to environmental
stresses via the Keap1-Nrf2-ARE pathway, Annu Rev Pharmacol Toxicol, 47 (2007)
89-116.
[56] M.A. Burguillos, T. Deierborg, E. Kavanagh, A. Persson, N. Hajji, A. Garcia-
Quintanilla, J. Cano, P. Brundin, E. Englund, J.L. Venero, B. Joseph, Caspase
signalling controls microglia activation and neurotoxicity, Nature, 472 (2011) 319-
324.
[57] S. Boillee, K. Yamanaka, C.S. Lobsiger, N.G. Copeland, N.A. Jenkins, G. Kassiotis,
G. Kollias, D.W. Cleveland, Onset and progression in inherited ALS determined by
motor neurons and microglia, Science, 312 (2006) 1389-1392.
[58] K. Yamanaka, S.J. Chun, S. Boillee, N. Fujimori-Tonou, H. Yamashita, D.H.
Gutmann, R. Takahashi, H. Misawa, D.W. Cleveland, Astrocytes as determinants of 175
disease progression in inherited amyotrophic lateral sclerosis, Nature neuroscience,
11 (2008) 251-253.
[59] M. Nagai, D.B. Re, T. Nagata, A. Chalazonitis, T.M. Jessell, H. Wichterle, S.
Przedborski, Astrocytes expressing ALS-linked mutated SOD1 release factors
selectively toxic to motor neurons, Nature neuroscience, 10 (2007) 615-622.
176
Chapter 4
H63D HFE – a genetic modifier for risk of ALS: summary and future perspectives
4.1. Introduction and Summary
Amyotrophic lateral sclerosis (ALS), though characterized by “selective” motor neuron degeneration in the brain and spinal cord, is a very heterogeneous disease with variable clinical phenotypes. Molecular mechanisms underlying ALS are complex and multifactorial. Despite the identification of over 15 genes associated with familial amyotrophic lateral sclerosis, the etiology of sporadic ALS, which accounts for majority of the cases, remains unclear. I have provided, in this thesis, increasing evidence suggesting a positive association between H63D polymorphism in the HFE gene (H63D
HFE) and ALS. Therefore, the main objective of this thesis is to identify how H63D
HFE gene variant contributes to disease processes in ALS. In order to address our goal, I generated two animal models in this study.
First, utilizing a H67D knock-in mouse line I provide direct in vivo evidence that the brain is not protected from iron dyshomeostasis associated with H67D HFE
(homologous to human H63D). The presence of H67D HFE significantly alters brain iron management protein expression and creates an environment for oxidative stress. It is noteworthy that the consequences of loss of iron homeostasis are manifested at younger ages, whereas adaptive responses against iron-induced toxicity are present at older ages.
Therefore, H63D HFE alone does not cause the level of neuronal degeneration that could lead to an ALS phenotype. However, in the presence of other risk factors or disease conditions, such as ALS, H63D HFE creates a permissive milieu that promotes the 177
convergence of disease mechanisms leading to neuronal degeneration as I demonstrated in the second part of this thesis.
Second, by generating a double mutant mouse line (SOD1/H67D) carrying both
H67D HFE and SOD1(G93A), which causes ALS phenotype, I provide evidence that
H63D HFE is a genetic modifier of ALS. The presence of H67D HFE shortens survival and accelerates disease progression in an ALS mouse model. We also identified mechanisms underlying this phenotype and our study suggests that an accelerated disease progression in ALS is associated with increased oxidative stress, increased microglial activation and altered iron homeostasis. Collectively, these findings have important contributions to our current understanding of the neurological consequences of H63D
HFE. The animal models developed for my thesis have opened up many opportunities for understanding how H63D HFE impacts disease processes, not only in ALS but also in other neurodegenerative diseases. Potential future studies stemming from the findings of this thesis are briefly discussed below. Cellular consequences of H63D HFE demonstrated in this thesis and in previous studies are summarized in figure 4-1.
178
Figure 4-1
Figure 4-1. Cellular consequences of H63D HFE based on findings from this thesis and published studies from our laboratory [1-5]. H63D HFE alters iron homeostasis, increases oxidative stress, increases glutamate release and induces mild endoplasmic reticulum
(ER) stress. Each of above cellular disturbances associated with H63D HFE creates an environment supporting inflammation, excitotoxicity and apoptosis that can be tolerated by neuronal cells, and also induces adaptive responses to protect against neuronal death as demonstrated in chapter 2 [5]. Additional insults or disease conditions, such as ALS demonstrated in chapter 3 and indicated by a green arrow in the figure, together with
H63D HFE may overwhelm the adaptive responses and eventually lead to apoptotic neuronal death.
179
4.2. Activating the Nrf2 Signaling Pathway as a Potential Therapy for ALS
One of the important findings in this thesis is that an accelerated disease progression in ALS is associated with increased oxidative stress, which I suggested may be the result of decreased nuclear factor E2-related factor 2 (Nrf2) level (chapter 3). The
Nrf2 signaling pathway is a major cellular defense mechanism against oxidative stress.
Under normal physiological conditions, Kelch-like ECH-associated protein 1(Keap1) protein links Nrf2 to the cytoskeleton and this promotes Nrf2 degradation. Oxidative stress induces dissociation of Nrf2 from Keap1 and stabilizes Nrf2. Nrf2 then translocates to nucleus, where it binds to the antioxidant response element (ARE) on the promoters of genes encoding antioxidant enzymes. This Nrf2-ARE transactivation leads to coordinated activation of endogenous antioxidant enzymes, such as glutathione peroxidases, catalase, peroxiredoxins and quinone oxidoreductases [6, 7].
Induction of these antioxidant enzymes enhances cellular resistance to oxidative and environmental stressors; whereas the deficiency of Nrf2 enhances vulnerability to reactive oxygen species (ROS)-mediated damage. For example, H-ferritin, the major iron storage protein mainly enriched in neurons, is one of the antioxidant proteins downstream of the Nrf2 signaling. In response to oxidative stress, H-ferritin expression is induced in an iron-independent manner by the binding of Nrf2 to the ARE on the promoter region of
H-ferritin gene [7-9]. Because Nrf2 levels are decreased in double mutant and SOD1 mice, H-ferritin expression is expected to decrease in these mice. However, H-ferritin levels are not changed in these mice despite decreased Nrf2 and a significant neuronal loss. This suggests increased H-ferritin expression in surviving neurons. Given that iron status can influence H-ferritin expression independently from the Nrf2 signaling and iron 180
dysregulation occurs in double mutant mice, it is reasonable to speculate that induction of
H-ferritin expression in surviving neurons of double mutant mice may be an adaptive response to iron accumulation rather than oxidative stress. However, iron-induced H- ferritin expression may not be sufficient to alleviate oxidative stress because other antioxidant proteins downstream of the Nrf2 signaling may not be induced in response to increased oxidative stress due to decreased Nrf2 levels in these mice. Collectively, the lack of adaptive responses by Nrf2-induced antioxidant protein expression increases the vulnerability of oxidative stress and enhances neuronal damage as we observed in double mutant mice. The Nrf2/Keap1 signaling pathway is summarized in figure 4-2.
In this study, I made the novel observation that the presence of H67D HFE in addition to SOD1(G93A) further decreases Nrf2 in the spinal cord of double mutant
(SOD1/H67D) mice (chapter 3). The effect of H67D HFE on Nrf2 signaling pathway is not fully understood. H67D HFE is associated with oxidative stress, iron dysregulation
(chapter 2, [5]) and endoplasmic reticulum stress [4]; each of which is known as an Nrf2 signaling inducer. Indeed in the presence of H67D HFE alone (H67D knock-in mice),
Nrf2 expression in the brain was upregulated (chapter 2, [5]). Thus, one would also expect increased Nrf2 in the spinal cord of double mutant (SOD1/H67D) mice, which is opposite from what we have observed. One possibility is that in double mutant mice
H67D HFE and SOD1 mutant exert synergistic effects on Nrf2 signaling independently from the Keap-1 pathway. Previous studies have identified Keap-1-independent regulation of Nrf2 involving glycogen synthase kinase (GSK) - 3β. GSK-3β phosphorylates Fyn protein and phosphorylated Fyn protein translocates to nucleus, where it phosphorylates Nrf2 promoting nuclear export and degradation of Nrf2 [10-13]. 181
GSK-3β activity is increased in a SOD1 ALS rodent model [14]. Moreover, previously we demonstrated that expressing H63D HFE in neuroblastoma cells upregulates GSK-3β activity [15]. Although GSK-3β activity was not examined in double mutant mice, it is tempting to speculate that double mutant mice, which carry both H67D HFE and SOD1 mutations, will have increased GSK-3β activity; therefore, the nuclear export and degradation of Nrf2 will be more active in these mice. Eventually, this will lead to decreased total Nrf2 expression followed by elevated oxidative stress in double mutant mice. GSK-3β activity in the double mutant mice should be measured as a next step. The proposed mechanism on the effect of H67D HFE on the Nrf2 degradation is shown in figure 4-2.
Oxidative stress has been well-accepted as an important neuropathological hallmark in ALS. However, antioxidant therapy has not shown much benefit in ALS patients [16]. Exogenous antioxidant therapy in neurodegenerative diseases is limiting due to the requirement of high concentration because of the limited passage of antioxidants via the blood-brain-barrier and the narrow therapeutic window due to their toxicity at higher concentration [17]. Therefore, the alternative strategy to reduce the detrimental effect of oxidative stress is activating endogenous enzymes; one approach is to induce the Nrf2 signaling. During the past six years, we have begun to appreciate the connection between oxidative stress and the abnormal regulation of the Nrf2 pathway in
ALS and the neuroprotective role of the Nrf2 have gained more attention in neuroscience.
Moreover, findings from this study (chapter 3) suggest decreased Nrf2 is an important driver of the increased oxidative stress contributing to accelerated disease progression in 182
ALS. This finding further supports the argument that manipulation of Nrf2 signaling is an attractive target to alleviate oxidative stress and ameliorate disease in ALS.
There are two possible approaches to increase Nrf2 levels. First and the more favorable approach would be to induce the Nrf2 signaling pathway. Preclinical studies using Nrf2 activators have shown benefits in animal models of neurodegenerative diseases [18, 19]. Of all Nrf2 activators, synthetic triterpenoids (CDDO-EA and CDDO-
TFEA), which are analogues of oleanolic acid, are potent activators of Nrf2 signaling pathway. These activators are bioavailable and can cross the blood-brain-barrier [18].
One study to date has reported that these Nrf2 activators, triterpenoids (CDDOs), increase expression and nuclear translocation of Nrf2 as well as upregulate anti-oxidant, anti- inflammatory and mitochondria biogenesis-promoting genes in SOD1(G93A) mice. In the same study, CDDO treatment started at presymptomatic age extends survival; whereas, treatment started at symptomatic age, which is more relevant with respect to clinical studies, prolongs disease duration in SOD1 mice [20]. Although the neuroprotective role of Nrf2 activators in this study needs to be confirmed, these findings suggest that activation of Nrf2 signaling is the feasible therapeutic target for treatment of
ALS and support my postulation that the decrease in Nrf2 is a driver of the accelerated disease noted in my model.
A second possibility to increase Nrf2 signaling is to limit the Nrf2 degradation.
For example, if increased GSK-3β is responsible for decreased Nrf2 as I proposed above,
GSK-3β inhibitors may have therapeutic benefits by limiting Nrf2 degradation. However, kinase inhibitors are diverse and their actions can have off-target effects. Thus, the major challenge in generating kinase inhibitors is achieving specificity, for example selective 183
inhibitors toward GSK-3α or GSK-3β. Moreover, GSK-3 is multifunctional and is essential for a variety of neuronal functions. Over-inhibition of GSK-3 is associated with neurodegenerative-like effects [21]. Therefore, compounds, which can exert mild inhibition to restore GSK-3 activity down to the physiological level, are of great interest as a potential therapy for ALS. Given that mechanisms underlying ALS pathogenesis are multifactorial, my data support the notion that activation of Nrf2 signaling, which appears to simultaneously alleviate multiple toxic factors implicated in ALS such as oxidative stress, neuroinflammation and mitochondrial dysfunction, is a potential therapeutic target in ALS. 184
Figure 4-2
Figure 4-2. Schematic diagram of the Nrf2-Keap signaling, marked with dark grey arrows, and the Nrf2 degradation pathway, marked with orange arrows. Under normal physiological conditions (1), Keap1 interacts with Nrf2 to retain Nrf2 in the cytoplasm.
This interaction promotes Nrf2 degradation and causes low basal expressions of antioxidant enzymes. Oxidative stress or electrophilic stimuli (2) induce Nrf2 release from Keap1. Nrf2 then translocates to the nucleus where it binds to antioxidant response element (ARE) and this interaction induces expressions of endogenous antioxidant enzymes. The Nrf2 degradation pathway involves phosphorylation of Fyn protein by
GSK-3β. Phosphorylated Fyn protein translocates to nucleus where it phosphorylates
Nrf2 promoting the nuclear export and degradation of Nrf2. H63D HFE may induce Nrf2 degradation by upregulating GSK-3β activity. 185
4.3. Caspase Inhibitors Hold Promise for Therapeutic Benefits in ALS
Another important finding in this study is higher L-ferritin and caspase-3 levels in double mutant mice. L-ferritin, primarily enriched in microglia, is an iron storage protein.
Microglia increase L-ferritin expression in response to increased iron accumulation
(chapter 2, [5]). In addition to iron accumulation, proinflammatory cytokines and inflammation associated with activating microglia can also induce L-ferritin expression and release in these cells [22]. Activated microglia can also increase caspase-3 expression without dying [23]. Therefore, increased L-ferritin and caspase-3 in double mutant mice suggest that increased microglial activation could be associated with the accelerated disease in these mice.
Though neuroinflammation indicated by activated microglia is the pathological hallmark of ALS, anti-inflammatory agents, such as minocycline, show no benefits in
ALS patients [24]. Microgliosis and increased inflammatory responses are closely linked in ALS, and activated microglia enhances the production of reactive oxygen species and pro-inflammatory cytokines, which can promote motor neuron stress and injury [25].
Therefore, delivering anti-inflammatory compounds into the hostile environment of activated microglia, which are capable of releasing more reactive oxygen species and pro-inflammatory cytokines, may not be an effective therapy. Perhaps, it is more appropriate to deliver the therapeutic agent that can prevent microglial activation and the release of cytokines. In support of this idea, a recent study by Burguillos et al., [23] reported that caspases, including caspase-3, -7 and -8, signaling controls microglial toxicity in the central nervous system (CNS). LPS treatment triggers expression of caspase-8 and caspase-3 in activated microglia, which is accompanied by increased 186
formation of nitric oxide synthase (iNOS) and reactive oxygen species (ROS). Caspases inhibitors or caspases knockdown using siRNA diminish microglial toxicity, mitigate the release of pro-inflammatory cytokines, and also reduce neuronal toxicity. These results indicate that caspases signaling controls microglial activation [23] and also support my notion that increased caspase-3 levels in double mutant mice are associated with increased microglial activation (chapter 3). These findings together support an argument that caspases inhibitors, which can block microglial activation and the release of pro- inflammatory cytokines, could be neuroprotective in ALS patients.
A neuroprotective effect of caspases inhibitors, DEVD-fmk and IETD-fmk, has been demonstrated in the rat LPS model and MPTP-lesion rat model of Parkinson’s disease (PD; [23]). In the LPS model, co-injection of the caspase inhibitor DEVD-fmk with LPS into the rat substantia nigra prevents LPS-induced microglial activation as well as mitigates the release of pro-inflammatory cytokines including TNF-α and IL-1β, and iNOS. In the MPTP-lesion rat PD model, intranigral injection of IETD-fmk inhibits caspase activation as well as MPTP-induced microglial activation and reduces neuronal toxicity [23]. These findings indicate that inhibiting caspases is a feasible target and further support my proposed opinion that caspases inhibitors will hold promise for therapeutic benefits in ALS patients.
187
4.4. Hepcidin may be Neuroprotective in ALS
In my thesis I have shown, decreased transferrin receptor (TfR) and increased L- ferritin is present in double mutant and SOD1 mice. H-ferritin expression, an iron storage protein mainly present in neurons, is not changed in double mutant and SOD1 mice compared to control mice despite a significant neuronal loss. As discussed earlier, the loss of Nrf2 would expect to result in a decrease in H-ferritin expression. Thus, the most plausible explanation for the lack of a decrease in H-ferritin is that surviving neurons have increased H-ferritin levels as an adaptive response to neuronal iron accumulation.
Therefore, the “normal” levels of H-ferritin combined with the decrease in TfR and increased L-ferritin together suggest increased iron accumulation in these mice. Based on the protein expression and indices of oxidative stress, it appears the loss of iron homeostasis is more extensive in double mutant mice than SOD1 mice. Although it is well-accepted that iron dyshomeostasis and oxidative stress is a significant part of ALS pathogenesis, the impact at the cellular level is unclear at present.
In the central nervous system (CNS), microglia can take up and store more iron than other glial cells [26-30] and are considered to play an important role in maintaining brain iron homeostasis. By sequestering iron in response to altered iron homeostasis and inflammation, microglia protect neurons against iron-induced toxicity [5, 22, 29, 30].
Therefore, it is reasonable to speculate that insufficient iron sequestering by microglia in response to neuroinflammation may contribute to iron-induced neuronal damage in ALS.
Thus, the therapeutic agent that can increase iron sequestering by microglia; thereby decreasing excess iron available to induce neuronal damage, could be a potential neuroprotective approach for ALS. 188
Normally, inflammatory stimuli increase microglial iron sequestering by modulating the expression of hepcidin, an iron management protein. Studies have demonstrated that inflammatory cytokines, in particular interleukin (IL)-6 can activate hepcidin gene expression via the STAT3 signaling pathway (reviewed in [31]).
Moreover, iron accumulation in cells induces hepcidin expression via the bone morphogenetic protein (BMP)-6-SMAD signaling pathway activated by hemojuvelin and/or holotransferrin/HFE/transferrin receptor 2 (Tf/HFE/TfR2) multiplex [31-33].
However, molecular signaling pathways underlying hepcidin regulation by HFE and
TfR2 are not completely understood. The function of hepcidin is to promote internalization and degradation of ferroportin, an iron efflux protein and the net result is decreased iron release and increased iron retention [31-33]. Therefore, hepcidin can be a link between inflammation and iron homeostasis.
Because neuroinflammation and increased inflammatory cytokines are present in
ALS, one would expect that increased hepcidin expression should accompany inflammatory responses in ALS. However, there is no increase in expression of pro- hepcidin, a precursor of hepcidin [34] in ALS patients despite increased iron and increased inflammatory responses. Moreover, there is a lack of a correlation between hepcidin and IL-6 that further suggests an uncoupling of inflammatory responses and iron homeostasis in ALS [34]. In addition, ferroportin mRNA levels are increased in the spinal cords of an ALS animal model [35] suggesting decreased hepcidin levels in these mice.
Hepcidin is widely expressed in the brain and spinal cord [36-38]. At cellular levels, inflammatory stimuli can increase hepcidin expression in microglia and astrocytes 189
but not in neurons. Hepcidin secreted from microglia and astrocytes decreases ferroportin expression in neurons, astrocytes and microglia [39]. Though, all brain cell types can accumulate iron, microglia and astrocytes are more resistant to iron-induced oxidative stress [40]. Moreover, microglia are more effective at taking up and safely sequestering iron than neurons and astrocytes [28, 40]. During the same amount of time, microglia can sequester approximately eightfold and fivefold more iron than astrocytes and neurons
[30]. Therefore, I speculate that decreased hepcidin levels in ALS can cause insufficient iron sequestering by microglia in response to neuroinflammation that eventually leads to increased excess iron available to induce neuronal damage. Therefore, hepcidin treatment to limit iron efflux and increase iron retention within microglia by downregulating ferroportin expression, may be neuroprotective in ALS.
190
4.5. Applications to the Human Population
In this study, I provided in vivo evidence demonstrating that H67D HFE alters iron metabolism, which in turn causes elevated oxidative stress and increased gliosis along with increased caspase-3 levels, contributing to an accelerated disease in ALS.
These findings support an argument that H63D HFE is a disease modifier for ALS. This notion is further supported by the previous study demonstrating that serum ferritin, which is negatively correlated with survival in ALS patients [41], is further elevated in ALS patients carrying H63D HFE compared to those carrying wild-type HFE [34]. Altered iron homeostasis ([42], chapter 3) together with increased serum ferritin and transferrin saturation [41, 43] in ALS patients raises the question of whether methods to reduce iron burden in ALS patients would have therapeutic values, particularly for patients carrying
HFE mutations. There are currently two methods for iron reduction, phlebotomy and iron chelation therapy.
At present, there is no study evaluating the effect of therapeutic phlebotomy in
ALS preclinical models or in patient populations. The expectation would be for phlebotomy to be performed early in the disease process rather than later as patients become weaker and have nutritional problems. Given that blood loss due to therapeutic phlebotomy induces erythrocyte heme synthesis [44] and an increase in erythrocyte regeneration is associated with increased oxidative metabolic activity [45], phlebotomized treatment may increase burden on energy demands in ALS patients. Even at the early stage of disease, ALS patients are hypermetabolic, on average they have a
10% higher resting energy expenditure than the control group [46]. Bouteloup et al., [47] also reported that hypermetabolism due to increased energy expenditure is present in 191
ALS patients early in the course of the disease. Moreover, Dupuis et al., [48] reported that defective energy homeostasis including increased lipid, protein and carbohydrate metabolism due to hypermetabolism is present early in the asymptomatic stage of the disease in an ALS model. Compensating increased energy expenditure with a fat- enriched high-energy diet reduces motor neuron loss and extends survival by 20% in an
ALS animal model [48]. Moreover, phlebotomy decreases triglyceride levels [49].
Hypermetabolism and lower serum triglyceride and cholesterol levels are negatively correlated with disease duration [47], survival [50] and respiratory impairment in ALS patients [51]. Triglyceride treatment improves motor performance, protects motor neuron loss and modestly extends survival in an animal model [52]. Together, these findings suggest that hypermetabolism and decreased triglyceride and cholesterol levels influence disease processes in ALS. Therefore, phlebotomy, which is associated with increased metabolic activity and decreased serum triglyceride levels, may not be favorable for ALS patients.
As an alternative to phlebotomy, excess iron levels can be reduced using chemical iron chelators. The neuroprotective effect of chemical iron chelators has been demonstrated in preclinical models of neurodegenerative disease including ALS,
Alzheimer’s disease and Parkinson’s disease [53] as well as in patients with neurodegenerative diseases, such as Friedreich ataxia [54], aceruloplasminemia [55] and
Alzheimer’s disease [56]. Although side effects associated with iron chelators have been reported in minority of patients, dose reduction alleviates or eliminates the adverse effects of iron chelators [54, 57]. In ALS, iron chelators delay disease onset, reduces motor neuron degeneration and extends survival in preclinical animal models [35, 58]. 192
Therefore, these studies suggest that iron chelation therapy should be evaluated as a potential treatment of ALS, perhaps particularly for those individuals with the HFE gene variant.
One potential therapy involving iron manipulation that has evolved from our project is the use of hepcidin to decrease excess iron level. Although the effect of hepcidin in ALS has not been studied, it is tempting to speculate that the use of hepcidin would have greater advantage over chemical iron chelators because hepcidin is naturally available peptide hormone that can increase iron sequestering by microglia/macrophages and thereby limit iron that would be available to cause neuronal degeneration. Thus, I suggest it is important to evaluate the possible beneficial effect of hepcidin for ALS using preclinical models.
In addition to therapeutic approaches to decrease iron, our data suggest that intervention strategies in ALS patients that increase iron should be re-evaluated in the context of the HFE genotype. For example, most ALS patients receive enteral nutrition treatment due to dysphagia and enteral nutrition regimens can last from 30 to 1460 days, depending on the rate of disease progression [59]. Because the H63D gene variant is associated with accelerated disease progression in ALS (chapter 3), long-term enteral nutrition treatment may worsen iron overload and could exacerbate the disease in ALS patients carrying HFE mutations. Our data argue for evaluating the effect of enteral nutrition treatment on clinical phenotypes in ALS patients carrying HFE mutations.
Finally, my data support the concept that genotype assessment in ALS patients will help clinicians make recommendation for patients’ families on which clinical trial may be more beneficial for the patient. For example, because decreased Nrf2 levels 193
appear to be an important driver to increase oxidative stress causing an accelerated disease in ALS (chapter 3) and neuroprotective role of Nrf2 activators has been reported in preclinical models of ALS [20], patients may be recommended to enroll in future clinical trials for Nrf2 activators. However, we demonstrated that the presence of H67D
HFE further decreased Nrf2 levels (chapter 3); therefore, Nrf2 activators treatment may be less effective in patients carrying H63D HFE compared to those carrying wild-type
HFE and require higher doses or longer dosing regimens to achieve similar effects. The
ALS patients carrying HFE mutations may be recommended to enroll in future clinical trials for compounds that can inhibit microgliosis, such as caspase-3 inhibitor discussed in previous section, because H63D HFE is associated with more active microgliosis that could be associated with an accelerated disease in ALS (chapter 3). Given that H63D
HFE is present in as many as 30% of ALS patients and we demonstrated in this study that the presence of H67D HFE influences disease pathogenesis in ALS, it is strongly suggested that HFE genotyping should be identified when evaluating therapeutic interventions for ALS in future clinical trials.
194
4.6. Conclusions
Before the production of the H67D knock-in mouse model, much of our knowledge of the functional consequences of HFE mutations originated from in vitro analyses utilizing neuroblastoma cell lines stably transfected with HFE variants.
Therefore, the H67D mouse model has opened up many opportunities to further advance our knowledge of an “old” protein, H63D HFE, and its “novel” role in neuroscience in addition to its “established” role in hemochromatosis.
The findings from this study indicate that the presence of the H63D polymorphism in the HFE gene alters brain iron metabolism, and creates an environment that promotes oxidative stress and gliosis. H63D HFE is a genetic risk for ALS and a modifier of the disease. The presence of this gene variant accelerates disease progression as well as shortens the survival in an ALS mouse model. Mechanisms underlying this phenotype include increased oxidative stress, microglial activation and iron dyshomeostasis. My data suggest that the Nrf2 activators, caspase inhibitors and hepcidin treatment, targeting each of the above disease mechanisms, may hold promise for therapeutic benefits in ALS patients (Figure 4-3). The animal models presented in this study appear to be good preclinical models and may create many opportunities to examine the impact of H63D HFE on disease processes and treatment strategies.
195
Figure 4-3
Figure 4.3. My thesis uncovered three putative pathogenic processes that converge to promote neuronal death. For future studies I propose the hypothesis that increased oxidative stress and microglial activation associated with altered iron homeostasis underlie the accelerated disease progression in ALS. Targeting each of above mechanisms can be a potential therapy for ALS as proposed below:
1. Nrf2 activators may alleviate oxidative stress by inducing endogenous anti-oxidant
gene expressions.
2. Caspase-3 inhibitors may block microglial activation, limit pro-inflammatory
cytokines release and reduce neuronal toxicity.
3. Hepcidin may increase microglial iron sequestration by downregulating ferroportin;
therefore, it may reduce excess iron available to induce neuronal toxicity.
196
4.7. References
[1] S.Y. Lee, S.M. Patton, R.J. Henderson, J.R. Connor, Consequences of expressing
mutants of the hemochromatosis gene (HFE) into a human neuronal cell line lacking
endogenous HFE, FASEB J, 21 (2007) 564-576.
[2] R.M. Mitchell, S.Y. Lee, W.T. Randazzo, Z. Simmons, J.R. Connor, Influence of
HFE variants and cellular iron on monocyte chemoattractant protein-1, J
Neuroinflammation, 6 (2009) 6.
[3] R.M. Mitchell, S.Y. Lee, Z. Simmons, J.R. Connor, HFE polymorphisms affect
cellular glutamate regulation, Neurobiology of aging, 32 (2011) 1114-1123.
[4] Y. Liu, S.Y. Lee, E. Neely, W. Nandar, M. Moyo, Z. Simmons, J.R. Connor, Mutant
HFE H63D protein is associated with prolonged endoplasmic reticulum stress and
increased neuronal vulnerability, The Journal of biological chemistry, 286 (2011)
13161-13170.
[5] W. Nandar, E.B. Neely, E. Unger, J.R. Connor, A mutation in the HFE gene is
associated with altered brain iron profiles and increased oxidative stress in mice,
Biochimica et biophysica acta, 1832 (2013) 729-741.
[6] H.E. de Vries, M. Witte, D. Hondius, A.J. Rozemuller, B. Drukarch, J. Hoozemans, J.
van Horssen, Nrf2-induced antioxidant protection: a promising target to counteract
ROS-mediated damage in neurodegenerative disease?, Free Radic Biol Med, 45
(2008) 1375-1383.
[7] T.W. Kensler, N. Wakabayashi, S. Biswal, Cell survival responses to environmental
stresses via the Keap1-Nrf2-ARE pathway, Annu Rev Pharmacol Toxicol, 47 (2007)
89-116. 197
[8] Y. Tsuji, H. Ayaki, S.P. Whitman, C.S. Morrow, S.V. Torti, F.M. Torti, Coordinate
transcriptional and translational regulation of ferritin in response to oxidative stress,
Molecular and cellular biology, 20 (2000) 5818-5827.
[9] Y. Tsuji, JunD activates transcription of the human ferritin H gene through an
antioxidant response element during oxidative stress, Oncogene, 24 (2005) 7567-
7578.
[10] M. Salazar, A.I. Rojo, D. Velasco, R.M. de Sagarra, A. Cuadrado, Glycogen
synthase kinase-3beta inhibits the xenobiotic and antioxidant cell response by direct
phosphorylation and nuclear exclusion of the transcription factor Nrf2, The Journal of
biological chemistry, 281 (2006) 14841-14851.
[11] A.K. Jain, A.K. Jaiswal, GSK-3beta acts upstream of Fyn kinase in regulation of
nuclear export and degradation of NF-E2 related factor 2, The Journal of biological
chemistry, 282 (2007) 16502-16510.
[12] A.I. Rojo, M.R. Sagarra, A. Cuadrado, GSK-3beta down-regulates the transcription
factor Nrf2 after oxidant damage: relevance to exposure of neuronal cells to oxidative
stress, Journal of neurochemistry, 105 (2008) 192-202.
[13] H.K. Bryan, A. Olayanju, C.E. Goldring, B.K. Park, The Nrf2 cell defence pathway:
Keap1-dependent and -independent mechanisms of regulation, Biochemical
pharmacology, 85 (2013) 705-717.
[14] J.H. Hu, H. Zhang, R. Wagey, C. Krieger, S.L. Pelech, Protein kinase and protein
phosphatase expression in amyotrophic lateral sclerosis spinal cord, Journal of
neurochemistry, 85 (2003) 432-442. 198
[15] E.C. Hall, 2nd, S.Y. Lee, N. Mairuae, Z. Simmons, J.R. Connor, Expression of the
HFE allelic variant H63D in SH-SY5Y cells affects tau phosphorylation at serine
residues, Neurobiology of aging, 32 (2011) 1409-1419.
[16] L. Ferraiuolo, J. Kirby, A.J. Grierson, M. Sendtner, P.J. Shaw, Molecular pathways
of motor neuron injury in amyotrophic lateral sclerosis, Nat Rev Neurol, 7 (2011)
616-630.
[17] B. Moosmann, C. Behl, Antioxidants as treatment for neurodegenerative disorders,
Expert opinion on investigational drugs, 11 (2002) 1407-1435.
[18] S. Petri, S. Korner, M. Kiaei, Nrf2/ARE Signaling Pathway: Key Mediator in
Oxidative Stress and Potential Therapeutic Target in ALS, Neurol Res Int, 2012
(2012) 878030.
[19] N.A. Kaidery, R. Banerjee, L. Yang, N.A. Smirnova, D.M. Hushpulian, K.T. Liby,
C.R. Williams, M. Yamamoto, T.W. Kensler, R.R. Ratan, M.B. Sporn, M.F. Beal,
I.G. Gazaryan, B. Thomas, Targeting Nrf2-mediated gene transcription by extremely
potent synthetic triterpenoids attenuate dopaminergic neurotoxicity in the MPTP
mouse model of Parkinson's disease, Antioxidants & redox signaling, 18 (2013) 139-
157.
[20] A. Neymotin, N.Y. Calingasan, E. Wille, N. Naseri, S. Petri, M. Damiano, K.T.
Liby, R. Risingsong, M. Sporn, M.F. Beal, M. Kiaei, Neuroprotective effect of
Nrf2/ARE activators, CDDO ethylamide and CDDO trifluoroethylamide, in a mouse
model of amyotrophic lateral sclerosis, Free Radic Biol Med, 51 (2011) 88-96.
[21] S. Hu, A.N. Begum, M.R. Jones, M.S. Oh, W.K. Beech, B.H. Beech, F. Yang, P.
Chen, O.J. Ubeda, P.C. Kim, P. Davies, Q. Ma, G.M. Cole, S.A. Frautschy, GSK3 199
inhibitors show benefits in an Alzheimer's disease (AD) model of neurodegeneration
but adverse effects in control animals, Neurobiology of disease, 33 (2009) 193-206.
[22] K.I. Rathore, A. Redensek, S. David, Iron homeostasis in astrocytes and microglia is
differentially regulated by TNF-alpha and TGF-beta1, Glia, 60 (2012) 738-750.
[23] M.A. Burguillos, T. Deierborg, E. Kavanagh, A. Persson, N. Hajji, A. Garcia-
Quintanilla, J. Cano, P. Brundin, E. Englund, J.L. Venero, B. Joseph, Caspase
signalling controls microglia activation and neurotoxicity, Nature, 472 (2011) 319-
324.
[24] P.H. Gordon, D.H. Moore, R.G. Miller, J.M. Florence, J.L. Verheijde, C. Doorish,
J.F. Hilton, G.M. Spitalny, R.B. MacArthur, H. Mitsumoto, H.E. Neville, K. Boylan,
T. Mozaffar, J.M. Belsh, J. Ravits, R.S. Bedlack, M.C. Graves, L.F. McCluskey, R.J.
Barohn, R. Tandan, Efficacy of minocycline in patients with amyotrophic lateral
sclerosis: a phase III randomised trial, Lancet neurology, 6 (2007) 1045-1053.
[25] S.H. Appel, W. Zhao, D.R. Beers, J.S. Henkel, The microglial-motoneuron dialogue
in ALS, Acta Myol, 30 (2011) 4-8.
[26] X. Zhang, M. Haaf, B. Todorich, E. Grosstephan, H. Schieremberg, N. Surguladze,
J.R. Connor, Cytokine toxicity to oligodendrocyte precursors is mediated by iron,
Glia, 52 (2005) 199-208.
[27] X. Zhang, N. Surguladze, B. Slagle-Webb, A. Cozzi, J.R. Connor, Cellular iron
status influences the functional relationship between microglia and oligodendrocytes,
Glia, 54 (2006) 795-804. 200
[28] A. Takeda, A. Devenyi, J.R. Connor, Evidence for non-transferrin-mediated uptake
and release of iron and manganese in glial cell cultures from hypotransferrinemic
mice, Journal of neuroscience research, 51 (1998) 454-462.
[29] S. Oshiro, K. Kawamura, C. Zhang, T. Sone, M.S. Morioka, S. Kobayashi, K.
Nakajima, Microglia and astroglia prevent oxidative stress-induced neuronal cell
death: implications for aceruloplasminemia, Biochimica et biophysica acta, 1782
(2008) 109-117.
[30] G.M. Bishop, T.N. Dang, R. Dringen, S.R. Robinson, Accumulation of non-
transferrin-bound iron by neurons, astrocytes, and microglia, Neurotoxicity research,
19 (2011) 443-451.
[31] C.C. Sun, V. Vaja, J.L. Babitt, H.Y. Lin, Targeting the hepcidin-ferroportin axis to
develop new treatment strategies for anemia of chronic disease and anemia of
inflammation, American journal of hematology, 87 (2012) 392-400.
[32] R. Deicher, W.H. Horl, New insights into the regulation of iron homeostasis,
European journal of clinical investigation, 36 (2006) 301-309.
[33] T. Ganz, E. Nemeth, Hepcidin and iron homeostasis, Biochimica et biophysica acta,
1823 (2012) 1434-1443.
[34] R.M. Mitchell, Z. Simmons, J.L. Beard, H.E. Stephens, J.R. Connor, Plasma
biomarkers associated with ALS and their relationship to iron homeostasis, Muscle &
nerve, 42 (2010) 95-103.
[35] S.Y. Jeong, K.I. Rathore, K. Schulz, P. Ponka, P. Arosio, S. David, Dysregulation of
iron homeostasis in the CNS contributes to disease progression in a mouse model of 201
amyotrophic lateral sclerosis, The Journal of neuroscience : the official journal of the
Society for Neuroscience, 29 (2009) 610-619.
[36] S. Zechel, K. Huber-Wittmer, O. von Bohlen und Halbach, Distribution of the iron-
regulating protein hepcidin in the murine central nervous system, Journal of
neuroscience research, 84 (2006) 790-800.
[37] Q. Wang, F. Du, Z.M. Qian, X.H. Ge, L. Zhu, W.H. Yung, L. Yang, Y. Ke,
Lipopolysaccharide induces a significant increase in expression of iron regulatory
hormone hepcidin in the cortex and substantia nigra in rat brain, Endocrinology, 149
(2008) 3920-3925.
[38] S.M. Wang, L.J. Fu, X.L. Duan, D.R. Crooks, P. Yu, Z.M. Qian, X.J. Di, J. Li, T.A.
Rouault, Y.Z. Chang, Role of hepcidin in murine brain iron metabolism, Cell Mol
Life Sci, 67 (2010) 123-133.
[39] P. Urrutia, P. Aguirre, A. Esparza, V. Tapia, N.P. Mena, M. Arredondo, C.
Gonzalez-Billault, M.T. Nunez, Inflammation alters the expression of DMT1, FPN1
and hepcidin, and it causes iron accumulation in central nervous system cells, Journal
of neurochemistry, (2013).
[40] S. Oshiro, M. Kawahara, Y. Kuroda, C. Zhang, Y. Cai, S. Kitajima, M. Shirao, Glial
cells contribute more to iron and aluminum accumulation but are more resistant to
oxidative stress than neuronal cells, Biochimica et biophysica acta, 1502 (2000) 405-
414.
[41] Y. Nadjar, P. Gordon, P. Corcia, G. Bensimon, L. Pieroni, V. Meininger, F.
Salachas, Elevated serum ferritin is associated with reduced survival in amyotrophic
lateral sclerosis, PloS one, 7 (2012) e45034. 202
[42] S. Oshiro, M.S. Morioka, M. Kikuchi, Dysregulation of iron metabolism in
Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis, Adv
Pharmacol Sci, 2011 (2011) 378278.
[43] E.F. Goodall, M.S. Haque, K.E. Morrison, Increased serum ferritin levels in
amyotrophic lateral sclerosis (ALS) patients, Journal of neurology, 255 (2008) 1652-
1656.
[44] P. Harper, T. Hybinette, S. Thunell, Large phlebotomy in variegate porphyria,
Journal of internal medicine, 242 (1997) 255-259.
[45] A.P. Yastrebo, Connection between Erythropoiesis and Oxidative Metabolism of
Bone Marrow, B Exp Biol Med-Ussr, 71 (1971) 111-&.
[46] J.C. Desport, P.M. Preux, L. Magy, Y. Boirie, J.M. Vallat, B. Beaufrere, P.
Couratier, Factors correlated with hypermetabolism in patients with amyotrophic
lateral sclerosis, The American journal of clinical nutrition, 74 (2001) 328-334.
[47] C. Bouteloup, J.C. Desport, P. Clavelou, N. Guy, H. Derumeaux-Burel, A. Ferrier,
P. Couratier, Hypermetabolism in ALS patients: an early and persistent phenomenon,
Journal of neurology, 256 (2009) 1236-1242.
[48] L. Dupuis, H. Oudart, F. Rene, J.L. Gonzalez de Aguilar, J.P. Loeffler, Evidence for
defective energy homeostasis in amyotrophic lateral sclerosis: benefit of a high-
energy diet in a transgenic mouse model, Proceedings of the National Academy of
Sciences of the United States of America, 101 (2004) 11159-11164.
[49] M. Kaito, M. Iwasa, Y. Kobayashi, N. Fujita, H. Tanaka, E.C. Gabazza, Y. Adachi,
Y. Kojima, N. Nakagawa, S. Watanabe, Iron reduction therapy by phlebotomy 203
reduces lipid peroxidation and oxidative stress in patients with chronic hepatitis C,
Journal of gastroenterology, 41 (2006) 921-922.
[50] J. Dorst, P. Kuhnlein, C. Hendrich, J. Kassubek, A.D. Sperfeld, A.C. Ludolph,
Patients with elevated triglyceride and cholesterol serum levels have a prolonged
survival in amyotrophic lateral sclerosis, Journal of neurology, 258 (2011) 613-617.
[51] A. Chio, A. Calvo, A. Ilardi, E. Cavallo, C. Moglia, R. Mutani, A. Palmo, R.
Galletti, K. Marinou, L. Papetti, G. Mora, Lower serum lipid levels are related to
respiratory impairment in patients with ALS, Neurology, 73 (2009) 1681-1685.
[52] W. Zhao, M. Varghese, P. Vempati, A. Dzhun, A. Cheng, J. Wang, D. Lange, A.
Bilski, I. Faravelli, G.M. Pasinetti, Caprylic triglyceride as a novel therapeutic
approach to effectively improve the performance and attenuate the symptoms due to
the motor neuron loss in ALS disease, PloS one, 7 (2012) e49191.
[53] X. Li, J. Jankovic, W. Le, Iron chelation and neuroprotection in neurodegenerative
diseases, J Neural Transm, 118 (2011) 473-477.
[54] N. Boddaert, K.H. Le Quan Sang, A. Rotig, A. Leroy-Willig, S. Gallet, F. Brunelle,
D. Sidi, J.C. Thalabard, A. Munnich, Z.I. Cabantchik, Selective iron chelation in
Friedreich ataxia: biologic and clinical implications, Blood, 110 (2007) 401-408.
[55] H. Miyajima, Y. Takahashi, T. Kamata, H. Shimizu, N. Sakai, J.D. Gitlin, Use of
desferrioxamine in the treatment of aceruloplasminemia, Annals of neurology, 41
(1997) 404-407.
[56] D.R. Crapper McLachlan, A.J. Dalton, T.P. Kruck, M.Y. Bell, W.L. Smith, W.
Kalow, D.F. Andrews, Intramuscular desferrioxamine in patients with Alzheimer's
disease, Lancet, 337 (1991) 1304-1308. 204
[57] G.L. Forni, M. Balocco, L. Cremonesi, G. Abbruzzese, R.C. Parodi, R. Marchese,
Regression of symptoms after selective iron chelation therapy in a case of
neurodegeneration with brain iron accumulation, Movement disorders : official
journal of the Movement Disorder Society, 23 (2008) 904-907.
[58] Q. Wang, X. Zhang, S. Chen, S. Zhang, M. Youdium, W. Le, Prevention of motor
neuron degeneration by novel iron chelators in SOD1(G93A) transgenic mice of
amyotrophic lateral sclerosis, Neuro-degenerative diseases, 8 (2011) 310-321.
[59] A. Molfino, I. Kushta, V. Tommasi, F. Rossi Fanelli, M. Muscaritoli, Amyotrophic
lateral sclerosis, enteral nutrition and the risk of iron overload, Journal of neurology,
256 (2009) 1015-1016. VITA
Wint Nandar
Education 2003-2007 Bachelor of Science, Biology, Radford University, Radford, VA, USA 2007-present Doctoral candidate, Neuroscience Graduate Program Pennsylvania State University, College of Medicine, Hershey, PA, USA
Awards 2013 Alumni Association Dissertation Award, Penn State University 2007 Dean Scholar of Biology Department, Radford University 2006 BBB National Biological Honor Society Undergraduate Research Award 2006 Virginia Academy of Science Best Student Paper Award 2006-2007 Dr. Fernande Aimee Juliette Gard Memorial Scholarship 2005-2006 Ricardo Grande Memorial International Scholarship
Selected Publications Nandar W, Neely EB, Simmons Z, Connor JR. H63D HFE accelerates disease progression in animal models of amyotrophic lateral sclerosis. Submitted to Brain. Nandar W, Neely EB, Unger EL, Connor JR. A mutation in the HFE gene is associated with altered brain iron profiles and increased oxidative stress in mice. Biochim Biophys Acta. 2013;1832: 729-741 Nandar W, Connor JR. HFE Gene Variants Affect Iron in the Brain. J Nutr. 2011;141: 729S-739S. Liu Y, Lee SY, Neely E, Nandar W, Moyo M, Simmons Z, Connor JR. Mutant HFE H63D Protein is Associated with Prolonged ER Stress and Increased Neuronal Vulnerability. J Biol Chem. 2011;286: 13161-13170. Hall EC 2nd, Lee SY, Simmons Z, Neely EB, Nandar W, Connor JR. Prolyl-peptidyl isomerase, Pin 1, phosphorylation is compromised in association with the expression of the HFE polymorphic allele, H63D. Biochem Biophys Acta. 2010;1802: 389-395. Nandar W, Milligan JM, Cline MA. Mechanisms of xenin-induced anorectic response in chicks (Gallus gallus). Gen. Comp. Endocrinol. 2008;157: 58-62. Cline MA, Nandar W, Hein PP, Denbow DM, Siegel PB. Differential feeding responses to central alpha-melanocyte stimulating hormone in genetically low and high body weight selected lines of chickens. Life Sci. 2008;83: 208-213. Cline M.A, Nandar W, Smith ML, Pittman BH, Kelly M, Rogers JO. Amylin causes anorexigenic effects via the hypothalamus and brain stem in chicks. Regul Pept. 2008;146: 140-146. Cline M.A, Nandar W, Prall BC, Bowden CN, Denbow DM. Central visfatin causes orexigenic effects in chicks. Behav Brain Res. 2008;186: 293-297. Cline MA, Nandar W, Rogers JO. Central neuropeptide FF reduces feed consumption and affects hypothalamic chemistry in chicks. Neuropeptides. 2007;41: 433-439. Cline MA, Nandar W, Rogers JO. Xenin reduces feed intake by activating the ventromedial hypothalamus and influences gastrointestinal transit rate in chicks. Behav Brain Res. 2007;179: 28-32.