Molecular neurology, Research Program Unit, Biomedicum 1 Neuroscience, Clinicum And Department of pathology, Medicum University of Helsinki Finland

GeneƟ cs of NeurodegeneraƟ on Alzheimer, Lewy body and motor neuron diseases in the Finnish populaƟ on

Terhi Peuralinna

Academic Dissertation

To be publicly discussed, with the permission of the Faculty of Medicine of the University of Helsinki, 29.8.2015 2pm in Meilahti srl Lecture Hall 4, Helsinki

Helsinki 2015 Supervisors Professor Pentti Tienari, MD, PhD Molecular neurology, Research Program Unit, Biomedicum 1 Neuroscience, Clinicum, University of Helsinki Department of Neurology, Helsinki University Central Hospital Helsinki, Finland

Docent Liisa Myllykangas, MD, PhD Department of pathology, Medicum University of Helsinki Folkhalsan Institute of Genetics Helsinki, Finland

Reviewers Docent Irma Järvelä Principal Investigator Department of Medical Genetics Institute of Biomedicine, Medicum University of Helsinki Helsinki, Finland

Professor Jari Koistinaho, MD, PhD Faculty of Health Sciences University of Eastern Finland, Kuopio, Finland

Opponent: Professor Henry Houlden Molecular Neuroscience, Institute of Neurology Faculty of Brain Sciences University College London London, UK

ISBN 978-951-51-1458-7 (paperback) ISSN 2342-3161 (print) ISBN 978-951-51-1459-4 (pdf) ISSN 2342-317X (online) http://ethesis.helsinki.fi /

Layout: Tinde Päivärinta, PSWFolders Oy

Hansaprint Vantaa 2015 In memory of my father

Author’s contact information

Terhi Peuralinna, MSc

Molecular Neurology research program Biomedicum Helsinki 1 Haartmaninkatu 8 00290 Helsinki Finland

Department of Pathology P.O.Box 21 (Haartmaninkatu 3) FI-00014 University of Helsinki Finland

Mobile Finland: +358 50 427 0394 Mobile USA: +1 404 406 2593 E-mail: [email protected] TABLE OF CONTENTS

ABBREVIATIONS LIST OF ORIGINAL PUBLICATIONS ABSTRACT 1. INTRODUCTION ...... 1 2. LITERATURE ...... 3 2.1 NEURODEGERATION ...... 3 2.2 ALZHEIMER’S DISEASE (AD) ...... 3 2.2.1 Defi nition of Alzheimer’s disease ...... 3 2.2.2 Epidemiology and clinical aspects of AD ...... 3 2.2.3 Pathology of AD ...... 3 2.2.4 Genetics of AD ...... 4 2.2.5 Amyloid Precursor Protein (APP) ...... 7 2.2.6 Hypotheses on pathogenesis ...... 7 2.3 CEREBRAL AMYLOID ANGIOPATHY (CAA) ...... 9 2.3.1 Defi nition ...... 9 2.3.2 Epidemiology and clinical aspects of CAA ...... 9 2.4 DEMENTIA WITH LEWY BODIES (DLB) ...... 10 2.4.1 Defi nition ...... 10 2.4.2 Epidemiology and clinical features ...... 10 2.4.3 Neuropathology ...... 11 2.4.4 Genetics of synucleinopathies ...... 12 2.4.5 Alpha-synuclein ...... 13 2.4.6 Hypotheses on pathogenesis ...... 14 2.5 AMYOTROPHIC LATERAL SCLEROSIS (ALS) ...... 15 2.5.1 Defi nition ...... 15 2.5.2 Epidemiology and Clinical aspects ...... 15 2.5.3 Genetics of ALS ...... 15 2.5.4 Hypotheses of pathogenesis ...... 16 2.6 GENOME ...... 17 2.6.1 Variations ...... 17 2.6.2 Single nucleotide polymorphisms (SNP)...... 18 2.7 GENE MAPPING STRATEGIES AND METHODS ...... 18 2.7.1 Linkage analysis ...... 18 2.7.2 Association analyses and Genome-wide association studies (GWAS) ...... 19 2.7.3 Candidate gene studies ...... 20 2.7.4 Polymerase chain reaction (PCR) ...... 20 2.7.5 Restriction Fragment Length Polymorphism (RLFP) ...... 21 2.7.6 Whole genome replication ...... 21 2.7.7 Sequencing ...... 22 2.7.8 TaqMan ...... 23 3. AIMS OF THE STUDY ...... 24 4. MATERIALS AND METHODS ...... 25 4.1 SUBJECTS ...... 25 4.1.1 Th e Vantaa 85+ study ...... 25 4.1.2 Th e Finnish ALS study ...... 26 4.1.3 Th e Cognitive Function and Ageing Studies (CFAS) ...... 26 4.2 Methods ...... 27 4.2.1 Genome wide genotyping ...... 28 4.2.2 Genotyping the SNCA polymorphisms ...... 28 4.2.3 Genotyping of the APP and APOE genes ...... 29 4.2.4 Statistical analyses and bioinformatics...... 29 4.2.5 Approvals for the studies ...... 29 5. RESULTS AND DISCUSSION ...... 31 5.1 Study 1 ...... 31 5.2 Study 2 ...... 32 5.3 Study 3 ...... 58 5.4 Study 4 ...... 35 6. CONCLUSIONS AND FUTURE PROSPECTS ...... 38 7. ACKNOWLEDGEMENTS ...... 39 8. REFERENCES ...... 40 ABBREVIATIONS

3R tau Th ree repeat tau protein/gene 4R tau Four repeat tau protein/gene Aβ Amyloid beta peptide ABCA7 ATP-Binding Cassette, sub-family A, member 7 AD Alzheimers Disease APOE ApolipoProtein E ALS Amyothorphic Lateral Sclerosis APP Amyloid Precursor Protein AβPP Amyloid-β Precursor Protein B3GALT4 Beta-1,3-Galactosyltransferase 4 BIN1 Bridging Integrator 1 c2orf73 Chromosome 2 Open Reading Frame 73 c9orf72 Chromosome 9 Open Reading Frame 72 CAA Cerebral Amyloid Angiopathy CADD Combined Annotation Dependent Depletion GARP Colgi-associated protein CASS4 Cas scaff olding protein family member 4 CD2AP CD2-Associated Protein CD33 CD33 molecule CELF1 CUGBP, Elav-like Family member 1 CERAD Consortium to Establish a Registry for Alzheimer’s disease CFAS Th e Cognitive Function and Ageing Studies CLU Clusterin CR1 Complement component (3b/4b) receptor 1 CSF CerebroSpinal Fluid ddNTP Di-DeoxyNucleotideTriPhosphate DLB Dementia with Lewy Bodies DNA DeoxyriboNucleic Acid EPHA1 Ephrin receptor A1 FALS Familial Amyothrophic Lateral Sclerosis FERMT2 fermitin family member 2 FTD Fronto Temporal Dementia GBA Glucocerebrosidase GWA Genome Wide Association GWAS Genome-wide association study HLA-DRB1/5 major histocompatibily comples, class II, DR beta1/5 Hsc70 heat shock chaperone 70 IFNK Interferon, Kappa INPP5D INositol PolyPhosphate-5-phosphatase LB Lewy Body LD Linkage Disequilibrium LN Lewy Neurites LRP Lewy Related Pathology LRRK2 Leucine-Rich Repeat kinase 2 MAF Minor Allele Frequency MEF2C Myocyte Enhancer Factor 2C MOBKL2B Mps One Binder kinase activator-like 2B MS4A Membrane-Spanning 4-domains, subfamily A NIA-RI National Institute on Aging and Reagan Institute NME8 NME/NM23 family member 8 NTP NucleotideTriPhosphate PCR Polymerase Chain Reaction PD Parkinson’s Disease PDD Parkinson’s Disease with Dementia PICALM Phosphatidylinositol Binding Clathrin Assembly protein PSEN1 PreSeniliini 1 PTK2B Protein Tyrosine Kinase 2 beta Q-Q plot Quantile-Quantile plot RFLP Restriction Fragment Length Polymorphism RNA RiboNucleic Acid RT-PCR Real Time Polymerase Chain Reaction SD Standard Deviation SLC24A4 Solute Carrier family 24, member 4 SNAP25 Soluble NSF Attachment Protein SNARE Soluble NSF Attachment protein receptor SNCA Alpha-Synuclein SOD1 Cu/Zn-superoxide dismutase 1 SORL1 Sortilin-Related receptor 1 TARPBP TAP binding protein TREM2 Triggering Receptor Expressed on Myeloid cells 2 TRIP4 Th yroid hormone Receptor Interactor 4 VPS52 Vacuolar protein sorting 52 homolog ZCWPW1 Zinc fi nger, CW type with PWWP domain 1 LIST OF ORIGINAL PUBLICATIONS

Th is thesis is based on the following original publications:

1. APOE and AßPP gene variation in cortical and cerebrovascular amyloid-ß pathology and Alzheimer’s disease: a population-based analysis. Peuralinna T, Tanskanen M, Mäkelä M, Polvikoski T, Paetau A, Kalimo H, Sulkava R, Hardy J, Lai SL, Arepalli S, Hernandez D, Traynor BJ, Singleton A, Tienari PJ, Myllykangas L. J Alzheimers Dis. 2011;26(2):377-85.

2. Neurofi brillary tau pathology modulated by genetic variation of alpha-synuclein. Peuralinna T, Oinas M, Polvikoski T, Paetau A, Sulkava R, Niinistö L, Kalimo H, Hernandez D, Hardy J, Singleton A, Tienari PJ, Myllykangas L. Ann Neurol. 2008 Sep;64(3):348-52. doi: 10.1002/ana.21446.

3. Chromosome 9p21 in amyotrophic lateral sclerosis in Finland: a genome-wide association study. Laaksovirta H, Peuralinna T, Schymick JC, Scholz SW, Lai SL, Myllykangas L, Sulkava R, Jansson L, Hernandez DG, Gibbs JR, Nalls MA, Heckerman D, Tienari PJ, Traynor BJ. Lancet Neurol. 2010 Oct;9(10):978-85.

4. Genome-wide association study of neocortical Lewy-related pathology. Terhi Peuralinna MSc1, Liisa Myllykangas MD PhD2,3, Minna Oinas MD PhD2,4, Mike A. Nalls PhD5, Hannah A.D. Keage PhD6,7, Veli-Matti Isoviita1, Miko Valori, Tuomo Polvikoski MD PhD8, Anders Paetau MD PhD2, Raimo Sulkava MD PhD9, Paul G. Ince MD10, Julia Zaccai, PhD7, George Savva, PhD7*, Carol Brayne MD FRCP FFPH7, Bryan J. Traynor MD MMSc MRCPI11, John Hardy PhD MD (Hon) FMedSci12, Andrew B. Singleton PhD5, Pentti J. Tienari MD PhD1,13. Accepted ABSTRACT

As the average age of death continues to rise in the whole world, the prevention and treatment of age- associated dementing neurodegenerative disorders will be important tasks to ensure better, longer and healthier life for the increasing elderly population. Alzheimer’s disease (AD) is the most common reason for dementia and is also the most common neurodegenerative disease. Th e second most common neurodegenerative disease is considered to be dementia with Lewy bodies (DLB) followed by Parkinson’s disease (PD). Amyotrophic Lateral Sclerosis (ALS) is a relatively common progressive neurodegenerative disease that aff ects motor nerve cells. Th is thesis project aimed to study the genetic background of common neurodegenerative diseases such as AD, DLB and ALS in the Finnish population. In the fi rst study, we analysed the role of the amyloid precursor protein (APP) and apolipoprotein E (APOE) genes in AD neuropathology. Mutations in the APP gene cause early-onset familial AD and there is some evidence that APP gene variants would play a role in late-onset AD, too. We genotyped >50 common variations in the APP gene and sequenced the APP promoter area to detect rare variations in the genes promoter area, both known and new. A few of these rare variations have previously been associated with AD. In the Finnish elderly population we found no APP variations to clearly associate with neuropathologically diagnosed AD or with any of the neuropathological features of AD such as cortical beta-amyloid, cerebrovascular beta-amyloid or neurofi brillary tangles. In addition, the APOEε4 results were updated, using the whole Vantaa85+ cohort. APOE ε4 is currently the strongest known risk factor for AD. We found a very strong association of APOE ε4 to all neuropathological features of AD. In the second study, we investigated how the common variants in the α-synuclein (SNCA) gene aff ect diff erent pathologies in the Vantaa85+ material. α-synuclein is the main component of the Lewy bodies, which are the pathological hallmarks of PD and DLB, and which are sometimes found also in AD. We genotyped 11 SNPs from the SNCA gene region. We found no association with the cortical beta-amyloid and only a faint one with Lewy related pathology. However, we found an association with neurofi brillary tangles (Braak stages IV-VI compared to 0-II). Th e associating SNP was rs2572324 with a p=0.004 aft er the Bonferroni correction. Th e two-locus analysis suggest an independent eff ect of APOE ε4 (OR=8.67) and rs2572324 (OR=3.36). In the subjects with both risk factors the eff ect was almost multiplicatively increased (OR=23.3). 35 subjects out of 38 with both risk factors had severe tau-pathology. Th ese results demonstrated the fi rst evidence for a role of genetic variation in SNCA in tau pathology. Moreover, beta-amyloid pathology was not associated with the SNCA variants, demonstrating a dissection of genetic eff ect on the two principal pathological features of AD. In the third study, we performed a whole-genome genotyping using Finnish ALS samples. Th e Vantaa85+ study subjects were used as controls. Th is study was the fi rst ALS genome-wide association study to fi nd a genome-wide signifi cant association. Th e location we found to be associated with familial ALS was on the chromosome 9p21, which had been noticed before, but the area had been much larger. Th e area has also been associated with frontotemporal dementia (FTD). FTD and ALS have been found to occur in same families. Now we managed to defi ne it with a 42 SNP haplotype. Th is considerably reduced the area of interest (down to 232 kb) and increased the possibility to fi nd the mutation behind the disease. Th is haplotype was found in around 40% of the Finnish ALS cases and likely explains the high rate of ALS as well as FTD in Finnish population. In the fourth study, we performed a genome-wide association study of DLB in the Vantaa85+ cohort and found two novel areas to be associated with DLB: Th e c2p21 location with 9 SNP haplotype (p=5.2x10-7) and the c6p21 location with 6 SNP haplotype (p=1.3x10-7). Th e c6p21 was signifi cant at the genome-wide level. Th e c2orf21 haplotype has two genes on its area, c2orf72 and SPTBN1. SPTBN1 is the candidate gene since it encodes beta-spectrin, a component of Lewy bodies, while c2orf72 is barely expressed in the brain. Th e c6p21 associated haplotype block is located in the HLA region and includes HLA-DPB1 and -DPA1 genes. In conclusion, these studies showed that the Finnish population is well-suited to study the genetics of neurodegeneration. We identifi ed genetic risk loci and variants for AD, DLB and ALS, some of which were previously known (APOE and chromosome 9 large region in ALS/FTD) and some were novel (SNCA in tau pathology, and DLB loci). Our results showed that neuropathologically defi ned parameters and diagnoses proved to be strongly associated with genetic risk factors, even with relatively few of samples. Hence, phenotypic precision (pathology) is an important element of the statistical “power” of a study. Introduction 1. INTRODUCTION

Th e average age of death continues to rise in the world. Th e number of deaths due to heart attacks, the main cause of death, has been declining since 1950 as the treatment, prevention and awareness has improved. Th e age of death is rising around the globe, and even though dementia risk is decreasing with improved education and better nutrition with rising age, the prevalence rate of dementia rises as well. Brain aging will have a substantial social impact since the fastest relative population growth in the western societies is in the oldest age-segments, as seen in Figure 1. Th e next huge step in this struggle for a better, longer and healthier, life is in the prevention of dementia and other neurodegenerative disorders. Alzheimer’s disease (AD) is the most common cause of dementia and also the most common neurodegenerative disease. Th e second most common neurodegenerative disease is considered to be dementia with Lewy bodies (DLB) followed by the Parkinson’s disease(PD). Age is the major predisposing factor for neurodegenerative diseases. Age is also a risk factor for the vascular catastrophes that lead to brain cell death in strokes. Th e knowledge of neurodegenerative diseases has greatly expanded during the last 30-40 years. Although Alzheimer’s disease was described more than 100 years ago (1906), it wasn’t until the early 1970s that it was recognized as a cause of most patients dementia. Parkinson’s disease was described even earlier in 1817 by British doctor James Parkinson. Dementia with Lewy bodies was described by the Japanese psychiatrist and neuropathologist Kenji Kosaka in 1976, but it wasn’t until mid-1990s that the disease started to be diagnosed.

Figure 1. Estimate of the population structure in year 2080 compared to the structure in 2013. Th is fi gure is provided by Eurostat (part of European Commission)(Eurostat 2014). Th e reproduction of the material produced by Eurostat is generally authorized by the European Union.

1 Introduction

Genetic research has had a major role in uncovering the molecular pathways involved in these disorders. Particularly identifying the gene defects of familial forms of disease have revealed crucial molecular players. For example fi nding the APP and alpha-synuclein mutations underlying the familial AD and PD led to the recognition of the major role of proteins coded by these genes in their pathogenesis. Later, multiplications of both APP and alpha-synuclein were shown to result in AD and PD. Th is confi rmed that overproduction of these proteins leads to aggregation and later to death of neurons. Th e genetic studies of common multifactorial forms of neurodegeneration have also provided insights into pathogenic mechanisms. Allele ε4 of apolipoprotein E, which has a role in lipometabolism, has been shown to be a major predisposing factor for AD. Recent genome-wide association studies (GWAS) have also revealed many variants of infl ammatory genes to be risk factors for neurodegenative diseases. Th is thesis project aimed to study the genetic background of common neurodegenerative diseases such as Alzheimer’s disease (AD), Cerebral amyloid angiopathy (CAA), Dementia with Lewy bodies (DLB) and Amyotrophic Lateral Sclerosis (ALS) in the Finnish population.

2 Literature 2. LITERATURE 2.1 NEURODEGERATION Neurodegeneration is a term for progressive loss of function and structure of neurons, usually leading to the death of neurons. Th e term neurodegeneration typically includes nerve cell death resulting from glial cell dysfunction, but excludes acute/subacute nerve death due to stroke, trauma, chemicals or infection. Depending on the cause and the location of the neuron loss the clinical symptoms vary. One molecular feature is common to many neurodegenerative diseases. Th ey usually involve aggregation of specifi c proteins inside and/or outside of the neurons. Neurodegenerative diseases can be grouped by the main aggregating protein. 2.2 ALZHEIMER’S DISEASE (AD) 2.2.1 Defi niƟ on of Alzheimer’s disease Alzheimer’s disease (AD) is the most common neurodegenerative disorder as well as the most common reason for dementia. Pathologically AD is characterized by senile plaques (SP) which are extracellular accumulations of Aβ peptide and by neurofi brillary tangles (NFT) which are intraneuronal accumulations of tau protein (Duyckaerts and Dickson 2011). AD is classifi ed as early-onset (<65 years ICD-10 code G30.0) and late-onset (>65 years ICD-10 code G30.1), oft en also called familial and sporadic, forms.

2.2.2 Epidemiology and clinical aspects of AD Th e prevalence of AD increases with age, so 1% of the age group of 65-69 years have AD where as 20-50% of people over 85 years have AD (Hy and Keller 2000). Th e incidence of AD dramatically increases with age. Th e estimates of the incidence of AD ranges from around 1 per 100 individuals per year in the age group of early 70s to the early 80s. Around late 70s or mid- 80s the incidence rate doubles to 2 per 100 individuals per year (Knopman 2011). Women have a higher risk of developing AD than men, even aft er correcting for the number of individuals at risk in each of the genders (Lautenschlager et al. 1996). Dementia has traditionally been the main characterization of clinically diagnosed AD. Early signs of AD are the disturbances in recent memory. Th e course of the disease is progressive.

2.2.3 Pathology of AD An Alzheimer’s brain is characterized with two pathologies: senile plaques and neurofi brillary tangles (NFT). Senile plaques are extracellular protein aggregates for which the main protein component is amyloid-β (Aβ). Aβ peptide is cleaved from the amyloid precursor protein (APP). Th e formation of Aβ is believed to be one of the critical steps that leads to a cascade of neurodegenerative events (Hardy and Selkoe 2002). Th e beta-amyloid structure is formed when these oligomers form fi brillary structures outside the neurons. Th ese formations are called amyloid plaques. Neuritic amyloid-containing plaques are associated with AD, whereas diff use plaques not containing the amyloid core are commonly found in normal aging. Neurofi brillary tangles are protein aggregates located inside neurons. Th e main components of NFT are hyperphosphorylated forms of protein tau. Tau protein normally interacts with tubulin to stabilize microtubules and promote tubulin assembly into microtubules. Upon hyperphosporylation tau loses its’ ability to bind to microtubuli, becomes mislocalised, and

3 Literature eventually forms neurofi brillary tangles. Six tau isoforms exist in brain tissue, and they are distinguished by their number of binding domains. Th ree isoforms have three binding domains (3R tau) and the other three have four binding domains (4R tau). All these six isoforms are found hyperphosphorylated in the neurofi brillary tangles (Iqbal et al. 2005). Th ere are several theories about what AD is and its pathology. Th e Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) defi nes pathological side of Alzheimer’s disease relying primarily to the density of amyloid plaques and secondary to the NFT (Fillenbaum et al. 2008; Mirra et al. 1991). In contrast the National Institute on Aging (NIA)-Reagan Institute (RI) criteria emphasises the role of NFT in the pathological defi nition AD, meaning that the both senile plaques and NFT should be found in the brain (Newell et al. 1999).

2.2.4 GeneƟ cs of AD In familial AD, which is also oft en called early-onset AD, the fi rst mutation was found in the gene coding amyloid precursor protein (APP) (Goate et al. 1991). APP protein is cleaved into several diff erent peptides and one of these is Aβ peptide. Aβ peptide is production is normal, but in the AD brain this peptide accumulates into extracellular beta-amyloid plaques. Several point mutations in the APP gene have been found to cause familial AD. One mutation in the APP (A673T) has been found to actually protect from AD (Jonsson et al. 2012). Mutations in two other genes have also been found to cause familial AD, presenilin 1 and 2 (Levy-Lahad et al. 1995; Rogaev et al. 1995; Sherrington et al. 1995). Presenilin 1 and 2 were subsequently identifi ed as components in the APP cleavage complex called gamma-secretase. Gamma-secretase generates the second cleavage and releases the amyloid-β peptide from APP as seen in Figure 2 (Wolfe et al. 1999).

Figure 2. Th e classic cleaving of the APP protein.

4 Literature

Genetics of sporadic AD is far more complex, but one gene has been found in all populations to aff ect the development of late-onset AD – Apolipoprotein E (ApoE). Th ere are 3 isoforms of ApoE: epsilon 2, 3 and 4. Th ey diff er in amino acid sequence of two residues: ApoE2 (cys112, cys158, ApoE3 (cys112, arg158), and ApoE4 (arg112, arg158). ApoE ε4 predisposes to AD, the net eff ect of the most common allele ApoE ε3 is considered neutral, although there is evidence that the ε3 can be divided into haplotypes that are either protective or predisposing (Myllykangas et al. 2002). ApoE ε2’s eff ect to AD is still under discussion, with some evidence that it is protective against AD (Corder et al. 1994). Th e genome wide association studies (GWAS) have found also several other genes involved in the risk of late-onset AD, indicating an important role of the immune system (CLU, CR1, ABCA7, MS4A cluster, CD33, TREM2 and EPHA1), cholesterol metabolism (APOE, CLU and ABCA7) as well as in the processes of cell membrane and synapse in late-onset AD (PICALM, BIN1, CD33, CD2AP, PTK2B and EPHA1) (Table 1). Th ese fi ndings have been verifi ed in several GWAS studies (Bertram et al. 2007; Lambert et al. 2013; Harold et al. 2009; Lambert et al. 2009; Hollingworth et al. 2011; Jonsson et al. 2012). Th e most recent predisposing mutations was found in the gene PLD3 (phospholipase D3), which have a role in processing of Aβ peptide (Cruchaga et al. 2014).

Table 1. Genes found and verifi ed in several genome-wide association studies to be predisposing to late-onset AD.

Acronym Gene Polymorphisms Function of the gene linked to AD CR1 Complement component rs3818361 Complement activation receptor (3b/4b) receptor 1 (immune system) MS4A Membrane-Spanning rs610932 Immune system cluster 4-domains, subfamily A rs2304933 rs4938933 CLU Clusterin rs11136000 Secreted chaperone protein (immune rs1532278 system, lipid metabolism) ABCA7 ATP-Binding Cassette, rs3764650 Possibly mediates cellular cholesterol sub-family A, member 7 and phospholipid release by apolipoproteins (lipid metabolism, immune system) (* EPHA1 Ephrin receptor A1 rs11767557 Part of immune system BIN1 Bridging Integrator 1 rs7561528 Nucleocytoplasmic adaptor protein, rs744373 synaptic vesicle endocytosis rs12989701 PICALM Phosphatidylinositol rs3851179 A clathrin assembly protein Binding Clathrin rs561655 Assembly protein CD33 CD33 molecule rs3865444 A member of the sialic acid-binding Ig-superfamily of lectins (immune system) CD2AP CD2-Associated Protein rs9349407 Protein regulates the actin cytoskeleton, endocytosis APOE Apolipoprotein E rs429358/C112R Lipid metabolism rs7412/C158R

5 Literature

Table 1 cont. Acronym Gene Polymorphisms Function of the gene linked to AD INPP5D inositol polyphosphate- rs35349669 Immune system, APP-metabolism 5-phosphatase MEF2C myocyte enhancer factor rs190982 Immune system and development of 2C muscles and nervous system. Protein possibly maintains the diff erentiated state of muscle cells HLA- major histocompatibily rs9271192 Protein is part of the heterodimer DRB1/5 comples, class II, DR presenting peptides from extracellular beta1/5 proteins. (immune system) NME8 NME/NM23 family rs2718058 Development of cilia and axonal member 8 transportation ZCWPW1 Zinc fi nger, CW type rs1476679 Epigenetic function with PWWP domain 1 PTK2B Protein Tyrosine Kinase rs28834970 Protein is involved in calcium- 2 beta induced regulation of ion channels and activation of the map kinase signaling pathway CELF1 CUGBP, Elav-like Family rs10838725 Regulates pre-mRNA alternative member 1 splicing and are possibly involved in mRNA editing and translation SORL1 sortilin-related receptor rs11218343 Involved in endocytosis, lipid LDLR class A repeats- transportation and APP-metabolism containing FERMT2 fermitin family member rs17125944 Involved in angiogenesis, adhesion 2 and tau-metabolism

SLC24A4 Solute Carrier family 24, rs10498633 a member of the potassium- member 4 dependent sodium/calcium exchanger protein family TRIP4 thyroid hormone rs74615166 Nucleus signaling, immune system receptor interactor 4

CASS4 Cas scaff olding protein rs7274581 Possibly regulates FAK and cell family member 4 spreading, associated with cancer

APP amyloid precursor rs63750847/ Cleaved to peptides like Aβ. Part of protein A673T lipid metabolism, signaling, peptides possibly bacteriocidal and fungicidal. TREM2 triggering receptor rs75932628/ A part of the immune response expressed on myeloid R47H and possibly involved in chronic cells 2 infl ammation

Data extracted from Morgan et al 2011 (Morgan 2011), Chouraki et all 2014 (Chouraki and Seshadri 2014), *Wu,C.A. et al 2013 (Wu, Wang, Zhao 2013)

6 Literature 2.2.5 Amyloid Precursor Protein (APP) Th e APP gene is located on chromosome 21 and contains at least 18 exons in 240 kilobases. Th e gene codes the amyloid precursor protein. Th e gene and protein has been named so because Aβ peptide, the major component of amyloid plaques, is cleaved from this protein. Th e gene was found in relation to AD and its cloning was based on the amino acid sequence of the Aβ (Kang et al. 1987). APP is a transmembrane protein which can be cleaved in two alternative pathways. Alpha-secretase pathway cleaves APP770, which is the largest APP isoform composed of 770 amino acids. APP770 is cleaved from the site 687 and this inhibits any formation of Aβ (Sisodia and St George-Hyslop 2002). Beta-secretase cleaves APP770 from 670 making it still possible for gamma-secretase to cleave APP from site 711 or 712 and this produces a 42-40 amino acids long protein called amyloid-β (Figure 2) (Sisodia and St George-Hyslop 2002). Th e function of the APP protein and its cleaved peptide fragments is revealing to be numerous and complex. Th e most substantiated role for APP could be in synaptic formation and repair, since APP expression is upregulated during neuronal diff erentiation and aft er neural injury. Th e cleaved intracellular domain of APP forms transcriptionally active complexes with the multidomain adaptor protein Fe65 and the histone acetyltransferase Tip60 (Cao and Sudhof 2001). Th ese complexes are consentrated in nuclear spots and are sites of active transcription (von Rotz et al. 2004). Th is indicates that APP has a role in gene expression. APP also has a link to the cholesterol balance. Th e alpha-secretase pathway stimulates cholesterol biosynthesis via the secreted ectodomain (sAPPα), whereas the beta-secretase pathway (sAPPβ) has the opposite eff ect (Wang et al. 2014).

2.2.6 Hypotheses on pathogenesis Th e mutations found to cause familial AD all seem to be part of the Aβ metabolism. Th ey usually enhance the production of Aβ or otherwise promote its oligo- and polymerisation. Amyloid plaques are created as the cell removes Aβ outside of the cell and forms more inert beta-sheet aggregates from the oligomers. Since the malfunction in amyloid cascade has been considered the main cause of the familial AD, it has also been the main target of the late onset AD research (Goate and Hardy 2012). Th e late-onset forms of AD seem to be caused by somewhat diff erent mechanisms (Figure 3). Th e Aβ metabolism is still important in the late-onset AD, but it is unknown if the amyloidosis is the primary reason for the disease or a by-product of other dysfunctions. Th ese dysfunctions may include disruptions in APP/Aβ metabolism, the cholesterol metabolism, in the synaptic and the membrane functions, as well as in immune responses. All these can contribute together with environmental factors to the development of late-onset AD (Chen et al. 2014). Evidence for the pathways where dysfunctions can happen have been obtained from the genetic fi ndings of late-onset AD (Table 1). APOE, besides its’ function as a lipid carrier, can bind to the Aβ and has a role in the degradation and clearance of the Aβ deposits by astrocytes (Koistinaho et al. 2004). Other genes such as CLU and CR1 reportedly have similar functions (Lambert et al. 2009). CLU and CR1 are also part of the complement system and thus could be part of the modulating immune system and infl ammation. CD33 and TREM2 seem to be involved in clearance of extracellular Aβ by the microglia (Lambert et al. 2013). Th ese fi ndings support the importance of the role of the immune system in the late-onset AD (Guerreiro and Hardy

7 Literature

Synaptic and Amyloid Cholesterol Immune Aβ metabolism membrane cascade metabolism response function

APP APOE APOE CLU PICALM PSEN1 CLU CLU CRE BIN1 PSEN2 ABCA7 ABCA7 ABCA7 CD33 INPP5D SORL1 MS4A CD2AP SORL1 CD33 EPHA1 APHA1 INPP5D CR1 PTK2B MEF2C SORL1 HLA-DRB1/5 SLC2A4 TRIP4 TREM2

Environmental factors

FAMILIAL AD SPORADIC LATE-ONSET AD

Figure 3. Th e pathways to the Alzheimer’s disease (Lambert et al. 2013; Morgan 2011).

2011) and suggest that neuroinfl ammation is part of the AD pathogenesis which contributes to neuronal damage (Bales et al. 2000). Some variations in genes like PICALM, BIN1, CD2AP and SORL1 have been found to be predisposing to AD. Th ese genes seem to be a part of the endocytosis functions, directing APP to degradation or to the amyloidogenic pathway. Th ey may also be part of regulating APP processing through cholesterol effl ux (Guerreiro and Hardy 2011; Chouraki and Seshadri 2014). Th ese fi ndings support the hypothesis that dysfunctions in Aβ clearance and degradation are important factors in the development of AD. Aβ metabolism has been noted to increase in oxidative and environmental stress situations. Environmental factors, such as repeated hits to the head, may have a role in development of AD. For example, boxers have an increased risk of developing AD type pathology (Jordan et al. 1997). Moreover, APP expression increases dramatically upon brain injury and is associated with Aβ deposition, especially with APOE ε4 positive subjects (Nicoll, Roberts, Graham 1995). Reoccurring or constant metabolic stress could also be caused by cardiovascular factors. Weakening vascular walls and slowed blood fl ow in the elderly could hinder the delivery of oxygen and nutrients as well as the removal of the waste from the brain. It has been shown that hypoxia increases both APP expression (Kalaria et al. 1993) as well as the activity of the beta- secretase pathway and Aβ production (Sun et al. 2006; Zhang et al. 2007). As the production of Aβ increases the degradation of the peptide becomes more important. Th e excess that can’t be degraded properly is removed from the cell for the microglia to clear. Failure in any of these steps

8 Literature could lead to the formation of the oligomers, which then increases the stress of the cell and again promotes Aβ production creating a vicious cycle. Formation of neurofi brillary tangles from the hyperphosphorylated tau is also an important part of the neuropathology of AD. Th e spreading of neurofi brillary tangles (Braak stage) is the critical parameter of AD pathology, which correlates with dementia. Genetic fi ndings support the role of tau in the disease (Chouraki and Seshadri 2014). BIN1 seems to be thus far the second most important genetic susceptibility locus. Th e BIN1 protein has been shown to interact with tau and modulate tau pathology (Chapuis et al. 2013). PICALM has also been found to be associated with neurofi brillary tangles and co-localized with abnormal tau (Ando et al. 2013). As the metabolism and clearance of the Aβ is taxing the cell, the clearance of other proteins might be hindered, leading to the accumulation of some other proteins like tau.

2.3 CEREBRAL AMYLOID ANGIOPATHY (CAA) 2.3.1 Defi niƟ on CAA is a disease of the blood vessels of the brain. In CAA a beta-amyloid similar to the one found to form plaques in AD brain parenchyme, forms fi brillary aggregates to the walls of the blood vessels (especially small arterioles) (Vinters 1987). Because of the aggregates, the blood vessel walls tend to weaken and crack. Th ese blood leaks in the brain lead to damages, like bleeding or hemorrhagic strokes (Vinters 1987).

2.3.2 Epidemiology and clinical aspects of CAA Th e deposits in CAA are very similar to the deposits found in AD brain, but having AD does not guarantee having CAA or vice versa. CAA is also found in asymptomatic old people and is actually quite common in old brains. A study by Attems and coworkers (Attems, Jellinger, Lintner 2005) found that 24% of dementia cases with AD pathology also had CAA pathology but 23.5% of the dementia cases having CAA did not have AD pathology. A population based study of old people (>85 years) made in Finland found that 69.6% of participants had CAA and it was more prevalent & severe in men than in women, but over all CAA was mild though prevalent among the elders (Tanskanen et al. 2012). CAA is also more common among the people with dementia (Tanskanen et al. 2013). Symptoms in CAA vary since the bleeding might happen in several diff erent parts of the brain (Vinters 1987). Some of the APP mutations that cause AD also results in severe CAA. Th ese mutations are located in the same region of the APP gene, right next to each other and hinder the degradation of the Aβ (Tsubuki, Takaki, Saido 2003). Th ese mutations are called the Dutch (E693Q), Arctic (E693G), Flemish (A692G) and Iowa (N694D) mutations, named aft er the place where the mutations were originally found (Grabowski et al. 2001; Kumar-Singh et al. 2002). Also the duplication of APP causes early onset AD with CAA (Sleegers et al. 2006). ApoE ε4 seems to predispose to CAA as well as to AD and the eff ect seems to be independent of AD (Attems, Jellinger, Lintner 2005; Tanskanen et al. 2005).

9 Literature 2.4 DEMENTIA WITH LEWY BODIES (DLB) 2.4.1 Defi niƟ on Dementia with Lewy bodies (DLB) is a progressive dementia syndrome associated with core neuropsychiatric features of cognitive function, visual hallucinations and parkinsonism (Ince 2011). Together with Parkinson’s disease with dementia (PDD) DLB is a part of a disease spectrum that associates with synucleinopathies (Spillantini and Goedert 2000). Even the name of PDD implies a patient fi rst diagnosed with Parkinson’s disease, who then develops dementia. Usually PDD diagnosis is recommended when PD develops fi rst, and dementia follows more than 1 year later. If the dementia develops fi rst or within one year of a diagnosis of PD, then a diagnosis of DLB is usually made (Ince 2011). DLB has also strong connections to AD, because there is also a Lewy-body variant of AD, meaning that the patient has both AD and DLB pathology. Th e co-occurrance of AD and DLB pathology seems to lead to a higher numbers of Lewy bodies and more severe dementia (Serby et al. 2003).

2.4.2 Epidemiology and clinical features DLB is the second most common neurodegenerative dementia syndrome in the elderly (Ince 2011). In autopsies in Finland 21% of over 70 year old patients with dementia have been found to have DLB like formations (Viramo and Sulkava 2010). In a Finnish population based autopsy study of over 85-year olds over 30% had limbic or cortical Lewy body formations (Oinas et al. 2009). Clinical features and diagnostic criteria are summarized in Table 2. Th e clinical characterization of DLB has been diffi cult, but certain “central”, “core” and “suggestive” features have been defi ned Table 2. DLB patients can have fl uctuating visual hallucinations and parkinsonisms. However the episodic memory is better in DLB than in AD. DLB is also very similar to Parkinson’s disease with dementia (PDD) (Noe et al. 2004). Th e main clinical diff erence between DLB and PDD is that in PDD the fi rst diagnosis is PD and the memory problems start later, usually years aft er a PD diagnosis.

10 Literature

Table 2. Clinical features and criteria needed for the diagnosis of probable and possible DLB (Modifi ed from McKeith et al. 2005)

CENTRAL FEATURE Dementia

CORE FEATURES Parkinsonism Fluctuating Cognition Visual hallucinations

SUGGESTIVE FEATURES Severe neuroleptic sensitivity REM sleep behavior disorder Low dopamine transporter uptake in basal ganglia (seen in SPECT or PET imaging)

SUPPORTIVE FEATURES Depression Delusions Hallucinations in other modalities Severe autonomic dysfunction Repeated falls and syncope Transient, unexplained loss of consciousness Relative preservation of medial temporal structures on seen in CT/MRI scan Generalized low uptake on SPECT/PET perfusion scan with reduced occipital activity Abnormal (low uptake) MIBG myocardial scintigraphy Prominent slow wave activity on EEG

COMMON SYMPTOMS (not required for diagnosis) Anxiety Apathy

DIAGNOSTIC CRITERIA: Probable DLB: • Dementia plus at least 2 Core features • Dementia plus one Core and at least 1 Suggestive feature Possible DLB: • Dementia and 1 Core feature • Dementia and at least 1 Suggestive feature

2.4.3 Neuropathology Protein aggregates called Lewy bodies and Lewy neurites are a major pathological feature of the DLB brain (Figure 4) (Spillantini et al. 1998). Alpha-synuclein (SNCA) is the main protein in these aggregates. Lewy bodies are round formations in the cell cytoplasm. Th ey can be devided in the classical brainstem Lewy bodies that have a dense center and a light halo around them, and cortical Lewy bodies that lack a halo. Lewy neurites constitute a string like formation that runs the length of a neuron’s axon and may exist in cells without classical Lewy bodies. Th ey are considered as the fi rst phase of the synuclein pathology. Together Lewy bodies and Lewy neurites are called Lewy-related pathology (LRP).

11 Literature

A consortium of international scientists known as the DLB Consortium recommends assessing alpha-synuclein in 10 brains areas for the neuropathological diagnosis of DLB. Th e widely accepted consept is that alpha-synuclein pathology spreads from the medulla into midbrain and then forebrain. Th e neocortical LRP is not only found in DLB, but also in PDD and Lewy body variant of AD, just as amyloid plaques occur in PDD and DLB cases (Kotzbauer et al. 2012; Lennox and Lowe 1997). DLB can be seen as the intermediate between the continuum from AD to PD (Lennox and Lowe 1997). Diseases in which LRP is the main cause of the neurodegeneration are called alpha-synucleinopathies (Spillantini and Goedert 2000).

Figure 4. Immunohistochemistry of the neocortex showing few Lewy bodies (long arrows) and Lewy neurites (short arrows). (courtesy of Dr. Minna Oinas)

2.4.4 GeneƟ cs of synucleinopathies Th e genetic background of pure DLB is not as known at the background of other common forms of neurodegeneration, such as AD. DLB has been considered to be a sporadic disorder, because only a few DLB families have been identifi ed and the disease is relatively late onset (Meeus et al. 2012). A twin investigation did not suggest a major genetic component in DLB (Wang et al. 2009). However genetic predisposition is supported by those few families in which a mixed phenotype of dementia and parkinsonism is inherited in an autosomal dominant or recessive manner (Meeus et al. 2012). Th e genes implicated in DLB are summarized in Table 3. Mutations in familial forms of DLB have been identifi ed in alpha-synuclein(SNCA), amyloid-β precursor protein (APP), and PSEN1 genes (Singleton et al. 2003; Singleton et al. 2002; Guyant-Marechal et al. 2007; Kaneko et al. 2007). With all of these mutations, the family has fi rst been diagnosed with either PD or AD, and then later on signs of DLB were noticed in some individuals in the family. Furthermore mutations in β-synuclein have been described (Ohtake et al. 2004) and the chromosomal area 2q35-q36

12 Literature has been found by linkage, although the gene and mutation are still unknown (Bogaerts et al. 2007). APOEε4, which has long been known to be predisposing to AD, reportedly predisposes to DLB as well (Kobayashi et al. 2011). Actually in synucleinopathies APOEε4 seems to increase the occurrence of dementia (Lamb et al. 1998; Tsuang et al. 2013). Heretozygous glucocerebrosidase (GBA) mutations, which as homozygous cause Gaucher’s disease, have been shown to increase the risk of DLB as well as PD (Mata et al. 2008). Th is elevated risk is possibly caused by the mutations in the GBA gene altering the structure of beta-glucocerebrosidase enzyme and thus impairing the function of lysosomes. As a result, alpha-synuclein may not be processed properly, allowing the formation of Lewy bodies (Cullen et al. 2011). Leucine-rich repeat kinase 2 (LRRK2) is a known PD gene, its protein is a part of Lewy bodies. Mutations in LRRK2 can cause familial autosomal dominant PD and predispose to PD or DLB (Kachergus et al. 2005; Ross et al. 2006).

Table 3. Genes associated to DLB

Gene Mutation Originally found FAMILIAL SNCA Triplication PD (Singleton et al. 2003) E46K, A53T and A53E (Polymeropoulos et al. 1997) (Figure 5 shows the locations of (Pasanen et al. 2014) the mutations) (Kruger et al. 1998) SNCB V70M and P123H DLB (Ohtake et al. 2004) APP Duplication AD (Guyant-Marechal et al. K670M, N671L and V717I 2007) PSEN1 ΔT440 AD (Kaneko et al. 2007) Locus 2q35-q36 Mutation and gene unknown DLB (Bogaerts et al. 2007) PREDISPOSING GBA N370S, L444P, D443N, D409V, Gaucher’s (Asselta et al. 2014) and D409H, and IVS10+1G>T* disease APOE Variant ε4 AD (Singleton et al. 2002) LRRK2 G2019S PD (Ross et al. 2006)

*Splicing mutation (in-frame exon-10 skipping)

2.4.5 Alpha-synuclein Th e Alpha-synuclein gene, SNCA, is located on chromosome 4q. Th e gene consists of 6 exons, which are spread over a 112 kb long segment. Th e gene’s mRNA product has at least 3 diff erent splicing variants known. Th e SNCA protein is mainly expressed in neurons and especially in presynaptic terminals. Th e protein is upregulated in a discrete population of presynaptic terminals of the brain during a period of acquisition-related synaptic rearrangement. SNCA probably has a chaperone function, since it functions with heat shock chaperone Hsc70 and SNAP25 to promote membrane SNARE complex assembly and exocytosis in neurotransmission. Th is also indicates a function in presynaptic signaling and membrane traffi cking (Lin and Farrer 2014) (Spillantini 2011).

13 Literature

Alpha-synuclein became an important focus of research aft er the fi rst autosomal dominant mutation linked to Parkinson’s disease was found. Th e fi rst mutation ever linked to familial PD was a point mutation in alpha-synuclein. Th is point mutation changed alanin at the position 53 to threonine (A53T) (Polymeropoulos et al. 1997). Finding how this mutation aff ected the cells was not easy since the point mutation in humans was actually the normal form of the alpha- synuclein in mice. Another model animal was needed and both drosophila and zebra fi sh models were developed (Feany and Bender 2000). With the fi nding of the SNCA triplication it was obvious that the overproduction of the alpha-synuclein was one of the reasons for the familial PD and DLB (Singleton et al. 2003). Th e fi rst mutation found in SNCA causing PD was shown to produce a more stable form of the protein, which hinders its degradation. Hence degradation of the protein is probably an important factor and might play a role in sporadic forms of DLB. Alpha-synuclein protein is usually soluble in physiological conditions and possibly even unfolded until it assumes α-helical conformation by binding lipids, but in synucleinopathies it forms toxic oligomers (Goedert 2001). Th ese oligomers then form insoluble fi brillary aggregations which constitute Lewy bodies. Th e aggregation mechanism is still unknown, and so is the structure of Lewy bodies, though they are thought to be a mix of unstructured, alpha-helix and beta-sheet rich conformers in equilibrium (Spillantini 2011).

Figure 5. Locations of the mutations in SNCA protein clustered in the loop. (Kara et al. 2013)

2.4.6 Hypotheses on pathogenesis Overexpression of α-synuclein, as happens especially in the SNCA triplication and duplication, aff ects cellular physiology, including mitochondria and proteasome functions, exocytosis, and protein biosynthesis, and it has been shown to induce the unfolded protein response and oxidative stress (Cookson and van der Brug 2008). Lewy bodies and Lewy neurites are known as well to inhibit proteolysis by proteasomes and increase the sensitivity of cells to a variety of toxic injuries such as mitochondrial damage (Dawson and Dawson 2003). It has been noted that patients with Gaucher’s disease (lysosomal storage condition) develop Lewy body pathology and patients with a heterozygous mutation in glucocerebrosidase (GBA) are susceptible to PD and DLB (Mata et al. 2008). Th ese fi ndings suggest that lysosomal dysfunctions might trigger the alpha-synuclein accumulation (Mata et al. 2008). Once the accumulation and oligomerization of the alpha-synuclein starts in one or several cells the pathology could possibly spread to neighboring cells and beyond in a prion-like manner (Brundin, Melki, Kopito 2010), however there is no evidence of these pathologies spreading from one subject to another (Beekes et al. 2014).

14 Literature 2.5 AMYOTROPHIC LATERAL SCLEROSIS (ALS) 2.5.1 Defi niƟ on ALS is a progressive neurodegenerative disease that aff ects nerve cells in the motor cortex, brainstem and spinal cord. As motor neurons – the neurons that control voluntary movements and muscle powe – degenerate, they can no longer send impulses to the muscle fi bres. When muscles are no longer receiving these impulses, the muscles become atrophic. A-myo-trophic comes from the Greek language. “A” means in Greek no or otherwise negative. “Myo” refers to muscle, and “Trophic” means nourishment. All these together the word then means “No muscle nourishment.” As a muscle has no nourishment it “atrophies” or wastes away. “Lateral” identifi es the areas in a person’s spinal cord where portions of the nerve cells that signal and control the muscles are located. As this area degenerates it leads to scarring or hardening (“sclerosis”) in the region. Th is term is slightly imprecise since it is currently known that ALS is a disorder of the whole motor neuron system including corticospinal tract, interneurons and lower motor neuron as well as the cortical bulbar tract. In the US ALS is usually referred as Lou Gehrig’s disease and in UK the commonly used term is Motor Neuron Disease (MND) (Strong et al. 2011). ALS has possibly become the most familiar term used for the disease since the famous “ALS Ice Bucket Challenge”, where you either pour a bucket of ice water on you or donate to ALS association, spread in 2014 in the social media to raise awareness and research funding for the disease (Siff erlin 2014).

2.5.2 Epidemiology and Clinical aspects Th e incidence of ALS worldwide is around 1.5 to 2.5 per 100000, though incidence is slightly higher in Finland as compared to most other European countries (Logroscino et al. 2010; Cronin, Hardiman, Traynor 2007). In Finland there are around 150-200 new ALS cases per year. Globally there are few high risk areas such as island of Guam and some parts of Japan. In these regions the disease is atypical and associated with dementia, parkinsonism and neurofi brillary tangles in many brain regions (Strong et al. 2011). Th e relatively selective degeneration of upper and lower motor neurons in ALS is associated with progressive wasting and weakness of skeletal muscles that leads to death from respiratory failure in the absence of respiratory support. Early symptoms of ALS oft en include increasing muscle weakness, especially involving the arms and legs, speech, swallowing or breathing. Usual age of onset is 50-60 years. Th e life expectancy of a classical ALS patient is 2 to 5 years aft er onset. Th e clinical diagnosis of ALS relies on the clinical picture (muscle weakness and waisting, fasciculations), electrophysiology (showing selective motor neuron degeneration with prominent fasciculations) and exclusion of other causes. Th ere are no specifi c cerebrospinal fl uid (CSF) biomarkers for ALS. Brain MRI is usually used to exclude other disorders, although sometimes T2 hyperintensities can be seen in the corticospinal or corticobulbar tracts. Gene tests are sometimes applied in the diagnostics of familial forms of ALS, especially in case of SOD1*D90A mutation (Andersen et al. 1995). Th is “Scandinavian” mutation, when homozygous, results in a slowly progressive form of ALS. Th e survival is typically >10 years.

2.5.3 GeneƟ cs of ALS ALS was fi rst described in the fi rst half of the 19th century and for decades it was thought to be a non-hereditary disease (Goetz 2000). In the 1950s the heritable factors were fi nally considered

15 Literature important in ALS etiology (Kurland, Mulder, Westlund 1955). Nowadays it is estimated that around 10% of the ALS cases are familial with a clear genetic heritance (Marangi and Traynor 2014). Familial ALS is usually dominantly inherited. Th e Cu/Zn-superoxide (SOD1) was the fi rst gene found to cause familial ALS (Rosen et al. 1993) and the mutations in this gene causes about 20% of the familial ALS cases. SOD1 mutations are also found in 2 to 3 % of apparently sporadic ALS cases (Logroscino et al. 2010). Th e most common mutation in SOD1 in Scandinavia is the D90A mutation which causes autosomal recessive slowly progressive leg-onset form of ALS. Th e D90A mutation can oft en be clinically distinguished from the other ALS cases even without genetic testing. Another possible characteristic of ALS-patients with SOD1 gene mutations might be a distinct metabolic profi le in the cerebrospinal fl uid (CSF) compared to other ALS patients (Kurland, Mulder, Westlund 1955), particularly in patients with the D90A mutation (Wuolikainen et al. 2012). Other genes that have been found and proven to be linked to ALS are TAR DNA binding Protein (TARDBP, protein called TDP-43) (Van Deerlin et al. 2008) and Fused with Sarcoma/Translocated in Liposarcoma (FUS/TLS) (Kwiatkowski et al. 2009; Vance et al. 2009). Several genes have been reported to be linked to ALS, but further evidence is still needed. Some variants which had been initially detected in ALS patients have later been found to be present in healthy controls, like with the Angiogenin (ANG). Th ere are also some concerns if all the mutations found in the SOD1 are actually pathogenic (Felbecker et al. 2010). Some of them might be just incidental fi ndings in aff ected individuals (such as SOD1*D90A heterozygocity). With many of these mutations there is a lack of evidence of segregation in families nor have they been seen in several aff ected individual and checked not to be found in normal individuals (Andersen 2006). Frontotemporal dementia (FTD) is known to appear together with ALS. Co-occurrence of dementia and motorneuron disease was noted already at 1993 (Mitsuyama 1993). Both ALS and FTD can occur in a family and the members can develop either one or both of the diseases. Th e location 9q21-q22 has been found to be linked to the diseases in such a family (Hosler et al. 2000). Another thing that links FTD and ALS is the accumulation of TDP-43 in both diseases as well as the notion that some mutations in TARBP gene can cause either ALS or FTD. As FTD can also be caused by mutations in MAPT, the gene coding tau-protein which aggregates in AD, ALS can be linked to the continuum of the dementia causing neurodegenerative diseases (Ng, Rademakers, Miller 2014).

2.5.4 Hypotheses of pathogenesis Th e cause of the process of neurodegeneration in ALS is mostly unknown, though there is evidence of glutamate excitotoxicity, infl ammation, mitochondrial dysfunction, apoptosis and proteosomal dysfunction having a role in the ALS pathogenesis. Th e only medicine that has thus far been found to slightly slow the progression of the disease in humans is riluzole which has anti-excitotoxic properties, but this medicine does not slow the disease process of all ALS patients (Sreedharan and Brown 2013). One thing that seems to be common with the genes related to familial ALS, is that they are oft en involved in RNA processing. TDP-43 is involved in regulating gene expression and RNA splicing (Ayala et al. 2005). In mutant form it accumulates in the neurons, is hyper-phosphorylated, ubiquitinated, and cleaved to generate C-terminal fragments (Neumann et al. 2007). Th is accumulation happens in both ALS and FTD cases. FUS/TLS protein binds to RNA and has functions in several processes. It is normally located in the nucleus. Th e mutant form accumulates in the neuronal cytoplasm

16 Literature

(Kwiatkowski et al. 2009). Th e SOD1’s most commonly known function is to be an antioxidant enzyme protecting the cell from reactive oxygen species toxicity (Vehvilainen, Koistinaho, Gundars 2014). However SOD1’s role in the neurons seems to be more diverse than expected and one of these functions possibly is in RNA binding (Bunton-Stasyshyn et al. 2014). Th ese genetic fi ndings indicate that the RNA processing is probably disrupted in the cells. Th e aggregation of proteins then would disrupt the cell functions even more. Th e TDP-43 and the SOD1 aggregates seem to spread to nearby neurons with prion-like manner, just like it seems to happen in the AD, DLB and the other neurodegenerative diseases with protein accumulation (Polymenidou and Cleveland 2011; Frost and Diamond 2010).

2.6 GENOME Th e genome is the complete set of an organisms DNA, including the DNA a in cell’s nucleus (3235Mb) as well as the DNA in the mitochondria (16,6kb). All the genes and everything between the genes consists of DNA and forms long strands. Th ese DNA strands form chromosomes together with proteins. Th e proteins and microRNAs together with the DNA and the external signals regulate which genes are used as a template for RNAs and how much of these RNAs are produced. Humans have 23 pairs of chromosomes and it is oft en said that the DNA is the blueprint of a person, like it is described in the Human Genome Project webpage (http://www.genome.gov/10001772). DNA could possibly be seen like a blueprint, but it is also the instructions for the use and repair. Like the blueprint and the instructions of a factory, if there are problems in the blue print then the factory might not be built well for its use, or maybe totally unable to function (birth defects and some miscarriages), if there are problems with the instructions the workers do not know how to do things properly (misfolding proteins, aggregation, hypermethylation etc.). Other factors also have a large impact, like the quality of the materials and the workers (the environment and mutations in genes).

2.6.1 VariaƟ ons Everyone has variations in their genome. Th ese can be single nucleotide polymorphisms (SNPs), insertion or deletions (indels), copy number variations, variable number tandem repeats, and large structural variations. Variable number tandem repeats are divided into microsatellites and minisatellites. Microsatellites repeat a sequence of 5 bp or shorter, where as minisatellites repeat a longer sequences. Th e variations make us diff erent from each other, though DNA sequence of two human beings is around 99.9% identical (Lander 2011). Sometimes these variations are actually useful for the individuals, like the coding mutation (A673T) in the APP gene that protects against Alzheimer’s disease and even against cognitive decline in the elderly who do not have Alzheimer’s disease (Jonsson et al. 2012). It is easier to notice the mutations that cause problems like the alpha-synuclein triplication which causes early-onset severe PD and/or DLB (Singleton et al. 2003). Most of the variation in the genome are non-functional and have little to no eff ect on the phenotype, but these variants can be used in tracking other variants that do have an eff ect through linkage disequilibrium (LD). LD is defi ned as an association of two linked alleles that happens more frequently than would be expected by chance. It refl ects the close vicinity of the alleles and correspondingly low probability of recombination separating the alleles (Bodmer 1972). When variations are used in this way, they are called genetic markers as the variation “marks” the

17 Literature neighbouring disease allele. Some of the fi rst disease association were reported using the ABO blood groups (chromosome 9) and transplantation antigens (HLA serotypes, chromosome 6) as genetic markers (Amiel 1967; Amiel 1967; Aird, Bentall, Roberts 1953). Possibly the fi rst disease locus detected by polymorphic DNA marker was reported in 1983 by Gusella et al. who found linkage between chromosome 4 and Huntington’s disease (Gusella et al. 1983).

2.6.2 Single nucleoƟ de polymorphisms (SNP) SNPs are the most common variation nowadays used as a genetic marker. Th e human genome is estimated to contain 10 million SNPs. Th e National Center for Biotechnology Information (NCBI) database contains data of 73 909 256 human SNPs with a verifi ed genotype. Th e frequency is known in 36 001 427 of these SNPs (data according to NCBI dbSNP Build 142, updated October 14th 2014) (http://www.ncbi.nlm.nih.gov/projects/SNP/snp_summary.cgi). Most of these are polymorphisms where one nucleotide has changed to another like an A to a T, but there are also one nucleotide indels meaning an insertion or a deletion of a single nucleotide. SNPs can be used to determine haplotype blocks. Th ese are regions of DNA that are usually inherited together and recombination at meiosis rarely brakes these areas. Th is oft en also means that the LD is strong between the SNPs in the same haplotype block. Population history shapes the haplotype blocks, in younger populations haplotype blocks tend to be larger. In populations that have been founded by a small number of individuals genetic variation is reduced. Th e populations that are both “young” and founded by relatively small number of people have long distance LDs, wide haplotype blocks and limited number of disease alleles. In such populations patients with a genetic disease are more likely to have the same mutation, whereas in more heterogeneous and older populations there is usually more diversity in the disease mutations. Long LDs also means that fewer genetic markers need to be genotyped to fi nd a marker that shows an association to the phenotype of interest, as one marker covers wider areas of the genome. Th e Finnish population is considered to be both relatively young and founded by relatively small number of people, and therefore genetic studies utilizing LD mapping of disease alleles may be especially well-suited to the Finnish population.

2.7 GENE MAPPING STRATEGIES AND METHODS Th e purpose of genetic mapping is the discovery of genes and variations that have an infl uence on the phenotype of interest. Usual variations used for gene mapping are SNPs and microsatellites. Microsatellites are oft en more informative since one marker has several alleles, but SNPs are more abundant and have higher density. Linkage and association analyses are strategies used in genetic mapping.

2.7.1 Linkage analysis Linkage analysis is an analysis that uses the recombination and the knowledge that the close areas in chromosomes tend to be inherited together. Following which markers are inherited together with the phenotype of interest will lead to the discovery of the location in the chromosomes where the gene aff ecting the phenotype is located. Large families are usually good for linkage studies as well as the knowledge of the pattern of inheritance. Usually the area found this way is quite large. Linkage analysis is a good method for simple Mendelian traits because there are only

18 Literature few models and they can be easily tested. Application to complex traits can be problematic since it may be hard to fi nd a precise model that adequately explains the inheritance pattern.

2.7.2 AssociaƟ on analyses and Genome-wide associaƟ on studies (GWAS) Association analyses study if the frequency of the genetic markers diff er between the subjects with phenotype of interest (cases) and those without the phenotype (controls). Th e higher the frequency is in the cases compared to those in the controls the more likely it is for the marker or something near it to have an infl uence to the phenotype in question. In GWAS the goal is to locate genes or other genomic elements that might have an eff ect to the phenotype of interest studying the whole genome of the participants. Th is is achieved with comparing single locus (usually SNP) allele/genotype frequencies and analyze whether they diff er between cases and controls. Th e GWAS era started in the second half of the last decade, when it became possible to genotype rapidly and with moderate costs several hundred SNP (Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research et al. 2007; Schymick et al. 2007; Rioux et al. 2007). With one SNP array it is possible currently to genotype up to 5 million SNPs depending on the chip used (www.illumina.com and www.aff ymetrix.com). Th e SNPs selected for genotyping on these chips are usually common and have relative high minor allele frequencies (MAF), but since these values diff er between the populations there are always some SNPs on an array that are not very informative (polymorphic) or the genotyping just fails. GWAS takes advantage of LD between SNPs and the variant contributing to the studied phenotype. However, found association might be a result of a false positive, badly designed study or just relevant for the population of the study and not found in other populations. Th e reason why the association might not be found in other populations could be that the actual mutation that causes the disease will be only found in certain populations. Another possibility is that the marker is for a multifactorial disease. In these diseases it is especially hard to determine the impact of possible associations since also the environment has an important role. Th e other population might not be as exposed to the environmental factors that are needed to trigger the eff ect of the variation. GWAS arrays, also known as beadchips, utilize the allele-specifi c primer extension. Th e DNA is fi rst amplifi ed with whole genome amplifi cation type of procedure (Page 39). Th en the DNA is fragmented to short strands. Th ese short strands are hybridized with the strands attached to the beads on the chip. Beads in diff erent locations on the chip have a diff erent DNA strand. Th e complement strands in the beads are either being a complete match to the genomic DNA strand or the last nucleotide on the chip strand is a mismatch. If it is a mismatch then in the exstaining, as the extend and stain phase is called, there is no extension and no detectable label (Figure 6). Th ose beads that have a strand that is all the way complement will have a label that is then detected by a machine. Aft er the beadchip is scanned and the available information from it gathered the data needs to be cleared. All the samples with low genotyping frequency, uninformative SNPs or which success rate is too low, and samples that are too closely related are removed.

19 Literature

Figure 6. Allele-Specifi c Primer Extension. Also used as a part of the whole genome arrays. Th e signal informs which allele is present.

2.7.3 Candidate gene studies Candidate gene studies use previous knowledge of the genes to determine which genes could possibly have an eff ect on the phenotype under research. Biological plausibility or previous research fi ndings are oft en the basis of candidate gene studies. Many diff erent genotyping methods are used in these targeted candidate gene analyses such as re-sequencing, Taqman, pyrosequencing, RFLP, etc. With next-generation sequencing methods whole-exome sequencing (WES) and whole- genome sequencing (WGS) are possible. Th ese sequencing methods are used to directly fi nd the variation causing the phenotype. Th e diffi culty with these methods is to fi nd and prove which of the mutations is pathogenic. One way is to use the candidate gene method and fi rst go through the genes which have previously been shown to aff ect the studied phenotype.

2.7.4 Polymerase chain reacƟ on (PCR) PCR can be considered the foundation of the modern gene technology on its own and it is also used in most of the protocols in genetic studies as a part of the procedure. Th e basic principle of the PCR is simple. DNA replication and repair is fundamental part of the function of the cells. When a cell divides the whole DNA of the cells needs to be replicated, this replication is done by the proteins called polymerases. Diff erent bacterial and eukaryote cells have slightly diff erent polymerases enabling some bacteria even to live in hot springs. Th e polymerases of these bacteria do not denature in hot temperatures. Th us these polymerases can be used in

20 Literature laboratories to replicate the DNA in vitro. All that is needed are: the template DNA, e.g. DNA separated from the blood cells; primer DNAs, which are short DNA strands that fl ank the area that has been targeted for replication; a supply of the four nucleotides that are the building blocks of the DNA; and the heat resistant polymerase. Heating these together over 95°C will separate the DNA strands of the template. Cooling the mixture to 50-72°C will let the primers to anneal with the template DNA. Th en the mixture is heated to 72-75°C for the optimal temperature for polymerase to copy DNA strand starting from the primers site. Th e mixture is heated back to 95°C to separate strands again and this whole cycle is repeated several times. Soon the mixture is mostly just copies of the target DNA sequence that is between the primers. PCR is used as a part of the methods such as the Sanger sequencing, pyrosequencing, RFLP and the TaqMan protocol.

2.7.5 RestricƟ on Fragment Length Polymorphism (RLFP) RFLP was the fi rst inexpensive method developed to genotype specifi c SNPs. Th e method uses restriction endonucleases to cut DNA segments within a specifi c nucleotide sequence. Prokaryote cells produce these diff erent enzymes to protect them from foreign DNA, like DNA from viruses. Th e nucleotide sequence that the enzyme cuts is called the restriction site, and even one nucleotide change in the site will prevent the enzymes function. In RFLP, the area around the restriction site is amplifi ed using PCR method before exposing it to the enzyme, aft er this the fragments are run on a gel electrophoresis, usually using an agarose gel, to separate diff erent sized DNA strands. Th ere is huge variation in the nucleotide sequences that are cut by the diff erent endonucleases, hence these can be used to genotype a huge variety of SNPs. Nowadays, with the rise of other faster, cheap and huge volume sequencing and genotyping methods, the use of RFLP has diminished.

2.7.6 Whole genome replicaƟ on Th e isotermal genome amplifi cation method can generate even 100kb long DNA fragments without sequence bias. Random hexamers bind to denatured DNA and are extended by a polymerase in an optimal temperature for the polymerase. As the polymerase moves along the DNA template it displaces the complementary strand which then becomes a template for replication (Figure 7). Th is method does not require diff erent temperatures like PCR, but the polymerase used should have 3’-5’ exonuclease proof-reading ability. Whole genome replication only works properly on genomic DNA, using a sample which has already been amplifi ed once would lead to shorter and shorter DNA strands with an increased risk of producing mutations. Th is method is one of the fi rst steps in the genome-wide association protocol, hence one should not use already amplifi ed DNA samples in GWA studies.

21 Literature

LongLong DNA fragmentsfragments (2-70(2-70 kbkb))

Figure 7. Whole genome replication

2.7.7 Sequencing Sequencing is a method to discover the nucleotide sequence of the target DNA. Sanger sequencing was developed by Frederic Sanger and colleagues in 1977 and is still in use with few modifi cations. Th e original method used radioactive di-deoxynucleotidetriphosphates (ddNTPs). DdNTP stops the DNA polymerase extending the DNA strand according to the complement strand. Th is way strands length will vary depending at which point the ddNTP will stop the DNA polymerase. Th e length of the strands indicates that in that location there is the nucleotide of which ddNTP the reaction was using. Th e method needs a PCR reaction for each of the four nucleotides. Th en these reactions are run on polyacrylamide-urea gel with a separate lane for a reaction, totaling four lanes (Sanger, Nicklen, Coulson 1977). Leroy Hood and his colleagues modifi ed Sanger’s technique and nowadays fl uorescently labeled ddNTPs are used and the phase aft er PCR is automated (Hood, Hunkapiller, Smith 1987). With this dye-terminator sequencing each of the four dideoxynucleotide chain terminators are labeled with a diff erent fl uorescent dye, which emits light at their own wavelength. Th is way only one reaction is needed to read the sequence of the targeted area. A machine is able to detect the diff erent wavelengths as the strands are separated by their size in the capillary electrophoresis.

22 Literature 2.7.8 TaqMan Th e basis of the TaqMan assay is the PCR method with one added ingredient, the TaqMan probe. Th is probe is a nucleotide probe that has a fl uorophore attached at the 5’-end and a quencher at the 3’-end. Th e quencher molecule quenches the fl uoroscence otherwise emitted by the fl uorophore. In order for the fl uorophore to emit it must be released from the probe. Th e TaqMan enzyme has a 5’ to 3’ exonuclease activity and will release the fl uorophore if it fi nds the probe on the area it is multiplying leading to the fl uoroscence to be emitted. But this happens only if the probe is totally annealed to its site on the template DNA from its 5’ -end. If there is even a nucleotide diff erence that prohibits the annealing of the fi rst nucleotide from the 5’-end, the enzyme will only push the probe from its way and continue, and there is no fl uoroscence signal. Th e TaqMan assay can be used to genotype SNPs as well as measure the amount of template DNA present, gene expression, viral load, bacterial identifi cation and is therefore widely used in research and clinical diagnostics.

23 Aims of the Study 3. AIMS OF THE STUDY

Th e general aim of the study was to uncover genetic factors underlining common neurodegenerative disorders.

Specifi c aims were: 1. To analyze the role of APOE and APP genes in Alzheimer’s pathological features in the population/based Vantaa85+ material. 2. To analyze the eff ect of common variation in the alpha-synuclein gene on amyloid-β, tau and lewy-related pathologies. 3. To analyze at the genome/wide level the genetic risk factors of ALS in the Finnish population. Th e control group constitutes the Vantaa85+ material, and genome-wide genome data is obtained for the subsequent studies. 4. To analyze at the genome-wide level the genetic risk factors of neocortical Lewy-related pathology (pathological correlated of dementia with Lewy-bodies)

24 Materials and Methods 4. MATERIALS AND METHODS 4.1 SUBJECTS 4.1.1 The Vantaa 85+ study Th e Vantaa 85+ study includes everyone over age 85 who were living in the city of Vantaa on April 1st, 1991. Of these 601 eligible subjects, 553 actually participated. 11 refused and the rest died before they could be examined. Th e clinical follow-up studies were made every 2 years until 2003. Of the 553 that participated, 306 consented to post mortem examination. Figure 8 gives an overview of the participants. Th is autopsy rate is worldwide the second highest among all population-based studies with neuropathological examination of the subjects (Zaccai, Ince, Brayne 2006). From the 553 participants there was viable DNA from 512 individuals and from the 306 autopsied there was viable DNA from 274 individuals. Th e neuropathological methods used in the Vantaa study are summarized in Table 4.

Figure 8. Flowchart of the Vantaa 85+ subjects

25 Materials and Methods

Table 4. Neuropathological analyses done in Vantaa 85+ study

Phenotypic feature Method Reference Cortical β-amyloid (Aβ) percentage Methenamine silver (Gundersen et al. 1988) (Polvikoski et al. ; Polvikoski et al. 1995) Cortical β-amyloid (Aβ) (Cerad score) Modifi ed Bielschowsky silver (Mirra et al. 1991) Neurofi brillary tau pathology (Braak Gallyas silver (Braak and Braak 1991) stage) Neuropathological AD Cerad score/Braak stage (Polvikoski et al. 2001) Cerebrovascular amyloid Congo red, (Tanskanen et al. 2005) immunohistochemistry (Peuralinna et al. 2011)

Brain stem and cortical Lewy-bodies Immunohistochemistry (McKeith et al. 2005) (Peuralinna et al. 2008)

Th e diagnosis of neuropathological AD was established according to the modifi ed NIA-RI criteria. Briefl y, diagnosis of AD according to the modifi ed NIA-RI criteria required amyloid plaque scores “moderate” or “frequent” according to the CERAD protocol and Braak stages IV– VI of neurofi brillary pathology (n = 113). Th e individuals belonging to the neuropathologically examined subpopulation who either lacked neocortical neuritic plaques or had the CERAD neuritic plaque score “sparse”, and whose Braak stage for neurofi brillary pathology was less than III, were defi ned as neuropathological controls (n = 48). Subjects of the neuropathologically examined subpopulation that did not fulfi ll the criteria for either of these categories were excluded. Th e severity of cerebrovascular A-β deposition was assessed by counting the percentage of the A-β immunoreactive blood vessels out of all vessels in four neocortical areas (frontal, temporal, parietal, occipital), hippocampus, and cerebellum to yield the individual percentage of the Aβ immunoreactive blood vessels (cerebrovascular Aβ deposition), throughout the whole section. Th e subjects with diff use cortical Lewy-related pathology were considered to have DLB. Th ere were 47 subjects with such pathology and they were used as cases. 194 subjects didn’t have any Lewy-related pathology (LPR) and were used as controls. Th ose who had Lewy-related pathology that was brainstem-predominant or limbic were excluded.

4.1.2 The Finnish ALS study Th e DNA used in this study was collected from patients attending an ALS specialty clinic at the Department of Neurology, Helsinki University Central Hospital that has received referrals from neurologists throughout Finland since 1994. Diagnoses were made according to the El Escorial criteria by a single neurologist, Dr. Hannu Laaksovirta. Th e study includes patients with familial ALS as well as those with sporadic ALS. Th e total amount of samples collected from diff erent patients was 442. Th e Vantaa 85+ study was used as controls for the Finnish ALS study.

4.1.3 The CogniƟ ve FuncƟ on and Ageing Studies (CFAS) Th e CFAS is a longitudinal, prospective, population-based cohort study of the United Kingdom’s Medical Research Council. It was initiated in 1989 and involves six medical centres in the UK 26 Materials and Methods

(www.cfas.ac.uk). Th e study includes a random sample of 18,226 people 65 years and over. A sub-sample of these people were asked whether they, with family support, were willing to donate their brain aft er their death. Th us far the median age at death for CFAS brain donors has been 87 years, but donations are still going on. A part of the CFAS sample set was used to verify the possible associations found in the Vantaa 85+ DLB GWAS. Only brains donated in two of the six centres participating in the CFAS study had their burden and anatomic distribution of alpha-synuclein investigated. Th ese were the only possible samples to use for verifying the fi ndings in the Vantaa 85+ GWAS. In July 2003 this included 208 investigated brains. Th e method to assess LRP has been previously described by Julia Zaccai and her colleagues (Zaccai, Ince, Brayne 2006). Th e current study included 54 subjects (cases) who showed LRP in at least one of the three regions brainstem, limbic or neocortex (brainstem only n = 24, limbic n = 2, brainstem+limbic n = 9, neocortical n = 19) and 138 subjects (controls) who did not show LRP in these regions. Th e controls included also the brainstem-negative amygdala- predominant group (n = 22) because subjects with this type of pathology were classifi ed as “control” in the Vantaa85+ material, since they were negative for brainstem and hippocampal LRP. Genotyping was successful in 51 cases with LRP and in 131 controls. Because of the lower sensitivity of the antibody used in the CFAS study, and lower number of subjects with neocortical LRP [19/208 (9%) overall; genotyping successful in 17 cases], the subjects with brainstem, limbic and neocortical Lewy-related pathologies were pooled together for the genetic analyses. Th us the number of cases was increased at the expense of the regional specifi city of LRP.

4.2 Methods Table 5. Summary of the methods used in the studies

METHOD REFERENCE EXPERIMENTAL PROCEDURES Polymerase Chain Reaction (PCR) (Kleppe et al. 1971) TaqMan® Assays-by-DesignSM Applied Biosystems, Foster City, CA Sanger Sequencing (ABI PRISM 3100) Applied Biosystems, Foster City, CA, US Whole genome amplifi cation Qiagen REPLI-g, Qiagen, Valencia, CA, US Genome wide SNP genotyping: Illumina Illumina, San Diago, CA, US Human370k &1M BeadChips RFLP ANALYSIS PROGRAMS Beadstudio 3.2 (genotyping module 3.2.32) Illumina, San Diago, US Sequencher 4.2 Gene Codes, Ann Arbor, MI, USA STATISTICAL METHODS AND SOFTWARE PLINK (Purcell et al. 2007) SPSS IBM, New York, US Haploview (Barrett et al. 2005) R (Team 2010) CADD (Kircher et al. 2014)

27 Materials and Methods 4.2.1 Genome wide genotyping Infi nium Human370 BeadChips (duo or quart)(Illumina, CA, USA), which assays 370 000 single nucleotide polymorphisms (SNPs) across the genome, was used for genotyping 873 samples from the Vantaa85+ and ALS studies. 110 samples were genotyped on Infi um Human 1M-Duo Beadchips. Genotyping was done according to the manufacturer’s recommendations. Reclustering was performed within the BeadStudio soft ware package (version 3.2, Illumina) using a no-call threshold of 0.15 again according to the manufacturer’s recommendations. Standard quality control procedures were applied as follows: exclusion of samples with SNP call rates of less than 95%, cryptic relatedness, non-European ancestry, minor allele frequency (MAF) less than 0.01, and Hardy-Weinberg equilibrium p value of less than 0.001. Non-European ancestry, or cryptic relatedness defi ned as identity-by-descent proportion of inheritance (pi_hat from the PLINK soft ware toolset version 1.06) 18 less than 0.1. Th e cryptic relatedness threshold led to the exclusion of individuals who share more than 10% of their genome, which meant that related individuals down to third or fourth degree relatives were not included in the fi nal analysis. All of the individuals were of European descent and none of the samples were excluded due to low call rates. Aft er cleaning the data in Vantaa 85+ included 218 subjects with 327 010 SNPs, including sex-chromosomal SNPs. Bonferroni corrected threshold for genome-wide signifi cance with this data is 1.56 x 10-7 (=0.05/327 010 SNPs). In the ALS study the sex-chromosomes were excluded leading for 318,167 SNPs passing the quality procedures. For genome-wide association analysis the study had the 405 cases and 497 controls since all the Vantaa 85+ samples could be used, not just those with pathological data. Th e area around the interesting peaks found in the GWAS studies were sequenced using the Big-Dye Terminator v3.1 sequencing kit (Applied Biosystems Inc., Foster City, CA, USA) and run on an ABI 3730xl genetic analyzer. Th e North American Brain Expression Consortium (NABEC) and United Kingdom Brain Expression Consortium (UKBEC) has gathered information of the brain mRNA expression and DNA methylation without determinable neuropathological evidence of a disease (Gibbs et al. 2010; Trabzuni et al. 2011; Trabzuni et al. 2012). Th is data is used to determine how genetic variability aff ects methylation and expression. Th e mRNA in these databases was assayed using Illumina HumanHT-12 v3 Expression Beadchips. Th e methylation was assayed on bisulfi te converted DNA using the Illumina Infi nium HumanMethylation27 BeadChips. Th e genotyping of the samples was performed using Illumina HumanHap550 v3, Human610-Quad v1 or Human660W-Quad v1 Infi nium Beadchips (Gibbs et al. 2010).

4.2.2 Genotyping the SNCA polymorphisms Genotyping of the SNCA markers was performed using whole-genome–amplifi ed DNA (Qiagen REPLI-g; Qiagen, Valencia, CA; www.qiagen.com). 11 single nucleotide polymorphisms (SNPs) within SNCA were genotyped using Taqman® Assays-by-DesignSM SNP Genotyping Services. Six of the SNCA SNPs were selected as tagging SNPs from the LD structure of the SNCA gene, and fi ve were selected from their association with PD. Th e locations of the SNPs and the method used are described in the article by Clarimon and colleagues (Clarimon et al. 2007). Th e genotype distributions did not deviate from Hardy–Weinberg equilibrium in the Vantaa 85+ material.

28 Materials and Methods 4.2.3 Genotyping of the APP and APOE genes A total of 60 SNPs were analyzed from APP gene and its fl anking region. 52 SNPs of those 60 analyzed were genotyped using Illumina HumanCNV370 Beadchips, 6 SNPs were genotyped with restriction endonuclease methods, and 2 SNPs using Sanger sequencing. Also the 971 bp region (−767 ± 204) of the APP promoter was sequenced in 40 subjects with very high Aβ % (≥7%) in order to fi nd previously reported rare mutations or novel variants. Th e region −90 ±204 was sequenced from whole study population in order to genotype two rare SNPs, the rs45476095 in the location −9 and the rs459543 in the location +37, as well as the novel variant at position −62 found in this study. Sequencing was performed using the Big-Dye Terminator v3.1 sequencing kit (Applied Biosystems Inc., Foster City, CA, USA), run on an ABI 3730xl genetic analyzer. Th e method for APOE genotyping has been previously described (Polvikoski et al. 1995).

4.2.4 StaƟ sƟ cal analyses and bioinformaƟ cs Sequence analyses in all studies were done with the Sequencer 4.5 Soft ware (Gene Codes Corporation, Ann Arbor, MI, USA). Th e associations between the APOEε4 and cortical and cerebrovascular A-β depositions were analyzed with PLINK (Purcell et al. 2007) (version 1.06), a toolset for genome association analysis using asymptotic Wald test. Th e association between age at death and cortical A-β deposition were analyzed by linear regression analysis using the SPSS for Windows v. 15.0. Association between APOEε4 allele and NIA-RI criteria based AD were also done with PLINK using Cochran-Armitage trend test. APP haplotype associations and LD structure were calculated with the Haploview soft ware, which uses an accelerated EM algorithm to estimate phased haplotypes based on the maximum likelihood determined from the unphased input. All other statistical analyses were performed using SPSS for Windows (v15.0). Whole genome association calculations were performed with the PLINK soft ware toolset (Purcell et al. 2007). P-values are based on the Cochrane-Armitage trend test. Odds ratios (OR) and upper and lower bound 95% confi dence intervals (CI) were computed for each SNP’s minor allele. Assuming a MAF of 0.15 and an adjusted level of signifi cance of ~2.0×10−7, the study had roughly 90% power to detect an associated SNP with odds ratios of around 2 under the additive model. Haplotypes were viewed and analyzed using Haploview 4.2. Manhattan plots were drawn on R using a gap package (Zhao 2007). QQ-plots were also drawn using R (R Development Core Team 2010). Th e population attributable risk percent (PAR%) was calculated using the formula: PAR% = p(r-1)/p(r-1)+1, where p is the proportion of the population exposed to the risk, and r is the relative risk. Th e possibility of the SNPs pathogenicity was discovered with the Combined Annotation Dependent Depletion (CADD) -tool (http://cadd.gs.washington.edu) (Kircher et al. 2014).

4.2.5 Approvals for the studies Th e Vantaa 85+ study was approved by the ethics committee of the City of Vantaa in 1991 (a consent was provided by the participant or their spouse or carer) and the medical ethics committee of the Hospital district of the Helsinki and Uusimaa updated the approval of the study in 2014 (74/13/03/00/2014). Further approval for the use of these samples for research purposes was obtained from the national board of medicolegal aff airs of Finland in 1996 and 2003 (DNro 2575/322/95).

29 Materials and Methods

Written informed consent for ALS genetic analysis was obtained from patients during their visit to the neurologist and this study was approved by the ethics committees for ophthalmology, otorhinolaryngology, neurology and neurosurgery (decision 68/2002) and internal medicine (Dnro 401/13/03/01/09) of the hospital district of Helsinki and Uusimaa, and by the insitutional review board of the National Institute of Aging (protocol number 2008-146).

30 Results and Discussion 5. RESULTS AND DISCUSSION 5.1 Study 1 APOEε4 has long been known to be a contributor to the development of AD. Having even one APOEε4 increases the risk of developing AD. Th is was also seen in the Vantaa 85+ population. Especially clear was the association with the amount of cortical Aβ found in the elderly brains with APOEε4 (p-value of 4.91x10-17). Th is is among the strongest reported association to the APOEε4 and refl ecting the power of a neuropathologically examined study. Th e amount of cerebral Aβ found in these brains ranged from 0% to 17.3%. Th e overall mean of the cortical Aβ accumulation in the brains of the very elderly was 3.41% (SD 3.8). But since there were those with very high Aβ accumulation and those with none, the median probably suited better to describe the overall Aβ accumulation in the brains of the elderly. Median accumulation was 2.31%. Th e mean cortical Aβ deposition in those with one or two alleles of APOE ε4 was 5.85% (SD=3.93, range 0%-15.66%) and in those lacking this variant (thus having APOE alleles either ε3 or ε2) the mean cortical Aβ deposition was 2.29% (SD=3.08, range 0%-17.31%). Th e APOE ε4 was also highly associated with the CAA phenotype (p-value=9.87x10-11). Th e overall mean of the CAA type Aβ accumulation was 4.13% (SD=8.6). When split with the APOE alleles the mean of the accumulation in subjects with one or two APOE ε4 alleles is 8.76% (SD=12.39, range 0%-72.67%), where as those with other APOE alleles the mean is only 2.00 (SD=4.82, range 0%-38.83%). As seen from the range of the values one can have an APOE ε4 allele and live a long life and die with hardly any detectable amyloid accumulation and those without APOE-ε4 can have a severe amyloid type pathology. Th ere was also no association between the age of death and the extent of cortical Aβ (p-value=0.52) or cerebrovascular amyloid-β (p-value=0.66). A clear association of APOE ε4 with tau-pathology was found. Using Braak stages grouped as 0-II versus IV-VI there was an association with a p-value of 3.45x10-7. All these associations were very robust and tell of the power of the Vantaa 85+ study even with the low number of participants. However, the extremely strong association between APOE ε4 and neocortical Aβ shows that it is possible and possibly even preferable to use quantitative traits in genetic studies. With a quantitative trait one does not need to decide which grade of pathology the subject has. Th e robust eff ect of APOE on neuropathology has also been shown in other population based neuropathologically examined studies (Nicoll et al. 2011). From the tau-pathology association one can see that dividing the population into cases and controls does also work. Part of this could be true since the population based and pathologically examined datasets are divided into cases and controls depending on their pathology. Th is creates controls who defi nitely do not have the particulate pathology in their brains and cases that only have the severe pathology and leaves those with intermediate pathology aside. So the controls are hyper normal, while the cases are in the other end of the neuropathologic spectrum. As both cases and controls are from the same area, population and age group this eliminates some of the environmental factors. APP has probably been the most studied gene regarding AD aft er the fi nding of the mutation in APP that causes familial AD. Th ere are several studies showing an association of the rare variants in the promoter sequence of the APP gene (Guyant-Marechal et al. 2007; Lv, Jia, Jia 2008; Nowotny et al. 2007). No association was found to the rare promoter variants or other variants in the APP gene to the late-onset AD or to either cortical or vascular Aβ deposits in the Vantaa85+ material. Th ere was one rare variant (rs11911934), which showed a very faint

31 Results and Discussion association (this can be seen in Table 6) but that was because only one subject had the variation. Th e person with the variation had NIA-RI criteria AD, hence the possible eff ect of this variation is inconclusive. Th e other rare variants genotyped were evenly distributed between cases and controls. Th e diff erence to the previous studies that have found an association to the rare variants in the promoter area might be that they have had an average age of onset under 70 years. Based on these fi ndings variations in the APP gene does not seem to predispose to the development of AD in the very elderly age group. One possible reason for this fi nding is that the people with these rare variations do not live as long as those who do not have them. Another thing to take into consideration is that these variations might be even rarer in the Finnish population and since the Vantaa85+ is quite a small data set, it does not have the power to detect their eff ect. Considering the Vantaa85+ age group, these are the ones who have survived to a very high age. Th us it would be maybe wiser to ask the question of what variation might be protective against diff erent pathologies, since there are those with very little or no pathologic accumulation of abnormal proteins.

Table 6. Summary of the results in Study 1.

Gene Pathology Variation p-values Findings APOE Cortical Aβ APOE ε4 4.91x10-17 Robust association to all pathologies, especially to Cerebrovascular Aβ APOE ε4 9.87x10-11 cortical A-β Neurofi brillary tangles APOE ε4 3.45x10-7

NIA-RI AD APOE ε4 1.62x10-8

APP NIA-RI AD rs459543 0.037 No association to the AD- related pathologies. Faint Cortical A-β N/A N/A association to NIA-RI AD which vanishes with Bonferroni correction. APP Cerebrovascular Aβ N/A N/A variations do not seem to Neurofi brillary tangles N/A N/A have an eff ect to the late- onset AD

5.2 Study 2 Th e alpha-synuclein is the major component in the Lewy bodies, but surprisingly there was only a very faint association between one SNCA variant and Lewy-related pathology, as can be seen in Table 7, and no association to amyloid pathology. SNPs genotyped later in the whole genome association study from the SNCA region did not show any association to the diff use cortical Lewy-related pathology either (Study 4). On the other hand several SNCA SNPs were associated with the neurofi brillary tangles (NFTs), one preserving the association even aft er Bonferroni correction. Th is SNP (rs2572324) had a p-value of 0.004 aft er the correction. Th is was also the SNP that faintly associated to the Lewy-related pathology Table 7. Th e rs2572324 genotype of T/T was overrepresented in cases where as genotypes C/T and C/C were underrepresented making the C-allele protective against NFT formation. Th ere has been evidence of an interaction between SNCA and tau proteins in vitro and in vivo, but this was now the fi rst time that an

32 Results and Discussion association between the SNCA gene and the tau-pathology was seen on the population level (Galpern and Lang 2006). SNCA might infl uence the aggregation of tau or possibly the expression of the tau protein. Since APOEε4 also associates to tau-pathology this raised a possibility that there is some interaction between these two loci or proteins. Th e odds rations (OR) were calculated for the diff erent situations. When the subjects had both the APOEε4 and the homozygous rs2572324 polymorphism T/T the OR for having tau-pathology was 23.3 (6.42-84.8). Th at is 35 subjects out of 38 had severe tau-pathology when they had both of these mutations. If the subjects only had APOEε4 the OR was 8.67 (3.09-24.3) and if they just had the homozygous SNCA polymorphism the OR was 3.36 (1.6-6.91). Clearly there is an eff ect with just the SNCA polymorphism making it an independent risk factor, and together these two factors seem to have a considerable eff ect. Wider and colleagues 2015 studied the SNCA polymorphism rs2572324 with samples from US, but did not replicate the fi nding (Wider et al. 2012). Th e reason for this could be the diff erent study design as the Wider et al study was not population based, had an admixed US population and diff erent pathologic evaluation.

Table 7. Summary of results in Study 2.

Sample set Pathology Variation p-values Findings Vantaa 85+ Neurofi brillary tangels rs2572324 0.0006 SNCA variations associate rs356165 0.009 with late-onset NFT pathology and rs258985 0.017 rs356168 0.045 rs258978 0.047 Neocortical Lewy- rs2572324 0.023 SNCA variation has a very related pathology faint association to late- onset neocortical LRP.

5.3 Study 3 Th e Finnish ALS study was the fi rst ALS GWAs to actually fi nd a genome-wide signifi cant association. Th e fi nding was a location associated with familial ALS. Th is location had been noticed before on the chromosome 9p21, but before the area had been much larger. Now it was defi ned to a 42 SNP haplotype. Th e 42 SNP haplotype covers 232 kb area that contains genes MOBKL2B, IFNK and C9orf72. Th e SNP with the lowest p-values were located between genes MOBKL2B and C9orf72 (Figure 9). Th is considerably reduced the area of interest and increased the possibility to fi nd the mutation behind the disease. As the Vantaa 85+ subjects were used as a control group the controls were not matched by age with the cases, but they were from the same population. Also the advantage of using the Vantaa85+ as controls was to have a control group that certainly did not have and did not develop ALS during their life. So again the controls can be considered as hyper normal compared to cases. Th is, together with Finnish population’s homogenous nature, possibly lead to enhancement of the strong association. Some of the ALS subjects involved in the study had the known recessive SOD1 mutation D90A and these subjects were also part of the whole genome genotyping to provide an association

33 Results and Discussion control. It was found that the 9p21 mutation is surprisingly common in Finland as 47.3% of the familial ALS cases had the mutation. Th e known SOD1 D90A-mutation was found only in 29.0% of the familial cases. So the 9p21 ALS mutation has clearly enriched in the bottle necks that have happened in the history of Finnish population and the high incidence of ALS in Finland can be at least partly explained with the enrichment of these mutations. Since the 9p21 locus has also been known to cause frontotemporal dementia (FTD), this mutation is possibly also behind the higher incidence of FTD in the Finnish population. All the exons and promoter areas from all the genes on the region were Sanger sequenced, but no point or other mutations were found to explain the cause of the areas clear association to the ALS. Th e results were published at this point to reveal the huge impact that this mutation has in the Finnish population and point to its relevance in ALS genetics (Table 8).

MOBKL2B IFNK

Figure 9. Th e chromosome 9 association results of 42 SNPs that form the haplotype associated with familial ALS. Th e SNP rs3849942 had the lowest p-value 4.15x10-13 and is highlighted.

Th e quest to fi nd the actual mutation was harder than thought. Finally, aft er more than 2 years of next-generation and Sanger sequencing the actual disease causing mutation was found in the gene c9orf 72 by two separate groups (Laaksovirta et al. 2010; Renton et al. 2011; DeJesus-Hernandez et al. 2011). Why was the mutation so hard to fi nd? Th e mutation was quite extraordinary for it was a massive expansion of a hexanucleotide repeat that consisted of only Gs and Cs. Th is repeat (GGGGCC) has thus far been un-PCRable and un-sequencable. Th is hexanucleotide expansion has now been found to exist around the world in diff erent populations and since the core haplotype is also found the mutation has probably derived from a single founder (Mok et al. 2012). Th e expansion is especially common in Northern Europeans, especially in Finns, but rare in Oriental populations (Majounie et al. 2012). Furthermore it has been found to be the most

34 Results and Discussion common genetic cause of ALS. Th e repeat is now been studied with huge interest as the mutation shows incomplete penetrance and varied phenotype. Th e C9orf72 protein function is still largely unknown and there are several hypotheses on how the repeat causes the diseases (Ash et al. 2013; Donnelly et al. 2013; Haeusler et al. 2014; Mori et al. 2013).

Table 8. Summary of the Study 3 results.

Sample set Variation p-values Findings ALS samples, Vantaa 85+ Chr 9 Previous chr9 fi nding defi ned to as controls rs3849942 9.1x10-11 42 SNP haplotype, C9orf72 as a 42 SNP haplotype 4.2x10-33 possible candidate gene.

Chr 21 (SOD1) SOD1 mutation in chr21 as rs13048019 2.6x10-8 positive control.

5.4 Study 4 In the fourth study a genome-wide association analysis was done of DLB in Vantaa85+. Th e cases (n=47) were defi ned as subjects who had diff use neocortical Lewy-related pathology (LRP) and the controls were subjects without any LRP in the brainstem and hippocampus (n=177). Anatomical spreading of LRP starts in the brainstem and neocortical LRP represent the most advanced stage of LRP. Hence the extreme phenotypes were used in the analyses, those with advanced LRP vs. those without any LRP. Such analysis has not been performed in a genome- wide study before. By analyzing 327 010 markers a few suggestive new loci were discovered. One locus provided genome-wide signifi cant signal (p=1.29 x 10-7, HLA-DPA1/DPB1) and 5 loci were found with suggestive association (p<10-5 on chromosomes 15q14, 2p21, 2q31, 18p11 and 5q23). Two of the 6 loci were marked by multiple markers, c2p21 (p=3.9 x 10-6), and HLA-DPA1/DPB1 (p=1.29 x 10-7). Th e location c2p21 contained a 9-SNP haplotype which associated with neocortical LRP, with SNP rs7595929 having the strongest association (Table 9). Th e c2p21 haplotype is located between two genes: c2orf73 and beta-spectrin family gene (SPTBN1). Th e haplotype block is downstream from c2orf73 and covers STBN1’s promoter. Th e whole C2orf73 gene and the SPTBN1 promoter and exons 1-3 (located within ~100 kb from the two top SNPs) were re-sequenced in cases with the risk haplotype. One missense mutation was found in the c2orf73 gene. Th is mutation was rs55714450 (Asn29His) and was weakly associated with DLB (p=4.2x10-3). However, this mutation does not seem to explain the association detected. Also expressions of the genes have to be considered and no mRNA expressions of c2orf73 was detected in frontal cortex samples with RT-PCR. Th is makes the c2orf73 an unlikely candidate gene.

35 Results and Discussion

SPTBN1, on the contrary to c2orf73, seems as an interesting candidate gene. Th ere is biological evidence of SPTBN1’s involvement in α-synucleinopathies. SPTBN1 protein has been shown to interact with alpha-synuclein both in vitro and in neuronal cells (Lee, Lee, Im 2012). Th e protein is also linked component of Lewy bodies (Lee, Lee, Im 2012). SPTBN1 is a good candidate gene and expression changes could have an eff ect in developing cortical LRP given that there were no variations resulting in amino acid changes in the SPTBN1 exons. Th e SPTBN1 gene is readily expressed and its protein has an important function in neurons. Th e SPTBN1 protein is cytoskeletal and forms heterodimers with α-spectrin. Th ese heterodimers then bind to other cytoskeletal proteins such as actin and ankyrin. Presynaptic SPTBN1/α-spectrin heterodimers have a role in stabilization of synapses and are involved with the regulation of exocytosis of neurotransmitters (Sikorski et al. 2000). Th ese heterodimers are also involved in the determination of cell shape, arrangement of transmembrane proteins, and organization of organelles (Sikorski et al. 1991; Sikorski et al. 2000). Six SNP haplotype is associated with cortical LRP on chromosome 6. Th e associated haplotype block is located in the HLA region at the location of HLA-DPB1 and -DPA1. One of the haplotypes is predisposing and another haplotype is protective. Th e subjects who were homozygous for the predisposing haplotype all also had homozygous HLA-DPB1*0201, whereas the subjects who were homozygous for the protective haplotype were carriers of HLA-DPB1*0401 (two homozygous, one heterozygous). HLA-DPB1/DPA1 area has previously been associated with immune system related disorders. Even the same type of predisposing and protection pattern has been reported in chronic beryllium disease (Fontenot et al. 2000). Association between PD and other HLA loci, HLA-DRA/DRB1 and HLA-DQB1, have also been reported (Hamza et al. 2010; Wissemann et al. 2013; Nalls et al. 2014). Th is may tell about the importance of antigen presentation and the immune system’s involvement in the development of LRP. HLA-DPB1/DPA1 locus variation may also have another function. In cis QTL analysis mRNA expression of vacuolar protein sorting 52 homolog (VPS52), TAP binding protein (TAPBP) and beta-1,3-galactosyltransferase 4 (B3GALT4) were infl uenced by the associated variatints. Th e methylation of the VPS52 was also aff ected by these variants. VPS52 is of special interest, because its’ yeast homologue is part of a Golgi-associated retrograde protein (GARP) complex (Conibear, Cleck, Stevens 2003) and GARP complex interacts with α-synuclein. Deletion of VSP52 disrupts GARP-complex and leads to vesicle aggregation induced by α-synuclein and causes toxicity in a yeast model (Soper et al. 2011). In order to replicate the fi ndings, two SNPs from the location chromosome 2 and fi ve chromosome 6 SNPs were genotyped in the CFAS material. In chromosome 2, one SNP associates with LRP, and haplotypes also associated with LRP in the CFAS study, one being predisposing and the other protective (Table 9). Th us chromosome 2 fi ndings were replicated in the CFAS material. However combined analysis of the SNPs did not reach genome-wide signifi cance. Th e joint analysis of the haplotypes almost reach genome-wide signifi cance (p=4.0x10-7). Th e chromosome 6 SNPs fi ndings were not replicated in the CFAS material as none of the SNPs signifi cantly associate with LRP in the CFAS material. However CFAS diff ers from the Vantaa 85+ material in several ways. Th e average age of the participants is somewhat lower than in the Vantaa 85+ and the techniques in LRP assessment were also diff erent using a lower sensitivity antibody (Alafuzoff et al. 2009). Th is all possibly contributes to the lower number of subjects with neocortical LRP in CFAS study. British population is also more heterogeneous than Finnish population making genetic associations harder to detect.

36 Results and Discussion

Th ere were also two SNPs, rs8041665 and rs8037309, which had the fourth and fi ft h strongest associations to neocortical LRP (both p=1.39*10-6). Th ey were located on chromosome 15, intergenic and only 259 bp apart. Th ese two SNPs were in total LD with each other in the Vantaa 85+ material. Th e CADD score (Kircher et al. 2014) suggests that the SNP rs8037309 possibly has a functional role, though its location is quite devoid of genes. Th e nearest gene is over 100kbp away from the SNP. However elements aff ecting gene regulation can be found at any distance (Shlyueva, Stampfel, Stark 2014). In conclusion, this GWAS on neocortical LRP has identifi ed suggestive novel genetic risk factors for predisposition of common late-onset neocortical LRP. Th e respective clinical disorder is DLB, which is the second most common neurodegenerative disease. Th is GWAS has unraveled novel molecules and pathways and will help researchers to design further studies for unraveling biomarkers and pathophysiologial mechanisms of this condition.

Table 9. Summary of the results of Study 4

Sample set Pathology Variation p-values Findings Vantaa 85+ Neocortical Lewy- 6p21 related pathology rs9277685 1.3x10-7 Haplotype: predisposing 9.5x10-8 protective 0.005 2p21 rs7595929 1.5x10-6 Variations in genes HLA- Haplotype 5.2 x10-7 DPA1/DPB1 and Spectrin possibly predisposes for late CFAS Neocortical Lewy- 6p21 onset- neocortical Lewy-related replication related pathology rs9277685 0.9685 pathology Haplotype 0.0639 2p21 rs4671212 0.0035 Haplotype predisposing 0.044 protective 0.011

37 Conclusions and Future Prospects 6. CONCLUSIONS AND FUTURE PROSPECTS

Finnish population proved to be well suited to study genetics of neurodegeneration with LD based methods. Th e genetic risk factor profi le of the Finns may be somewhat diff erent compared to other Caucasian populations, possibly due to genetic isolation of the Finns.

Specifi c conclusions: • Alpha-synuclein polymorphisms associate to tau pathology on the population level. • APP promoter or intronic polymorphisms do not seem to have an eff ect on the late- onset AD or any pathologies related to AD; cortical- or cerebrovascular amyloid and neurofi brillary tangles. • APOE is as in all populations prominent factor for the development of the late-onset AD. Th e clear association in the Vantaa85+ material reveals the power of the population based, pathologically examined studies. • Two new areas, potentially genes STBN1 and HLA-DPB1/DPA1, were found to be associated with DLB. • A familial mutation causing ALS was verifi ed to be located on chromosome 9p21 and was defi ned to a 42 SNP haplotype. Th is haplotype was unexpectedly common in Finland and may explain together with the SOD1 D90A-mutation the high incidence of ALS in Finnish population.

Th e molecular mysteries behind the cause of neurodegenerative diseases are unravelling slowly and mechanisms have shown to be complex. Th e disease processes seem more like a spectrum of protein aggregation with diff erent pathologies overlapping and infl uencing each other than separate diseases. From this point of view, studying the molecular aspects of the brains and comparing those to the genetic fi ndings is a sensible method. We should not only wonder why does our brains suff er from these diseases but also what makes one have AD instead of DLB, DLB instead of PDD, PDD instead of PD and so on. Since the ALS caused by c9orf72 hexanucleotide repeat has already proven to be very interesting with its incomplete penetrance and varied phenotype expression, the next step is to fi nd what are the factors involved in modulating the phenotype.

38 Acknowledgements ACKNOWLEDGEMENTS

Th is project has been carried out in Molecular Neurology Research program, Biomedicum, University of Helsinki, Finland and Laboratory of Neuropathology, Department of Pathology, University of Helsinki, Finland. Laboratory of Neurogenetics, National Institute on Aging, NIH, Bethesda, MD, USA

I want to thank all the people who have contributed to this work. My special gratitude to my supervisors Pentti Tienari and Liisa Myllykangas for believing in me and sticking with me to the end. For encouraging me to learn new methods and broaden my skills (I actually did enjoy participating in that autopsy). Th ank you to all my co-writers. Special thanks to Minna Oinas, Maarit Tanskanen, Anders Paetau and Tuomo Polvikoski for sharing their knowledge of neuropathology. Hannu Lehtovirta for introducing me to the world of ALS. Lilja Jansson for her skill in the laboratory (especially for extracting DNA from those thawed and again frozen ALS samples) and her willingness to teach those skills. Raimo Sulkava for starting the whole Vantaa85+ project and still being so enthusiastic about it. John Hardy for being so enthusiastic, inspirational, encouraging and even willing to share his offi ce. Everyone in the Neurogenetics laboratory on NIA for always making me feel welcome, especially Andrew Singleton (even if I’m your favourite employee just because you didn’t have to pay my salary), Dena Hernandez, Mike Nalls, Sampath Arepalli. Particularly people in Neuromuscular Diseases Research Group (Laboratory of Neurogenetics). Bryan Traynor for always being there to mentor and help me, even bringing food when I had a horrible fl u. Jennifer Schymick for being an awesome partner in the laboratory and a joy as a roommate. United College London Neurogenetics Laboratory staff , who helped and tolerated my two week sequencing marathon. Sonja Scholz for being a friend as well as a co-worker in both USA and UK. To all my friends in Finland and abroad especially Eili Huhtamo, who has been invaluable support, since the beginning of my genetic studies. Maija Arponen for editing my thesis as well as for her friendship and support, even travelling to Atlanta to help me. Mikko Pervilä, Nikodemus Siivola and Hanna Järvinen for their advise and just being there to listen to my troubles. My family for their love and support. Especially my husband Teemu Vilen for agreeing to be a single father of twins at times. Also to my mother for providing me home and help me whenever I have needed it. And my in-laws whose help have been crucial for last 4 years. My speech therapist Caitlin Solari for training and helping me to gain back some of my abilities. Th is work was fi nancially supported by the Helsinki University Central Hospital, the Samfundet Folkhälsan, the Academy of Finland, the Sigrid Juselius Foundation, the Finnish Parkinson Foundation, Microsoft Research Foundation, ALS Association and DPBM doctoral school. Also CCP Games, which provided offi ce space in Iceland as well as in Atlanta.

39 References REFERENCES

Aird I, Bentall HH, Roberts JA. 1953. A relationship between cancer of stomach and the ABO blood groups. Br Med J 1(4814):799-801. Alafuzoff I, Ince PG, Arzberger T, Al-Sarraj S, Bell J, Bodi I, Bogdanovic N, Bugiani O, Ferrer I, Gelpi E, et al. 2009. Staging/typing of lewy body related alpha-synuclein pathology: A study of the BrainNet europe consortium. Acta Neuropathol 117(6):635-52. Amiel JL. 1967. Study of the leukocyte phenotypes in hodgkin’s disease. In: Histocompatibility testing. Curtoni ES, Mattiuz PL,Tosi MR, editors. Copenhagen, Denmark: Munksgaard. 79 p. Andersen PM. 2006. Amyotrophic lateral sclerosis associated with mutations in the CuZn superoxide dismutase gene. Curr Neurol Neurosci Rep 6(1):37-46. Andersen PM, Nilsson P, Ala-Hurula V, Keranen ML, Tarvainen I, Haltia T, Nilsson L, Binzer M, Forsgren L, Marklund SL. 1995. Amyotrophic lateral sclerosis associated with homozygosity for an Asp90Ala mutation in CuZn-superoxide dismutase. Nat Genet 10(1):61-6. Ando K, Brion JP, Stygelbout V, Suain V, Authelet M, Dedecker R, Chanut A, Lacor P, Lavaur J, Sazdovitch V, et al. 2013. Clathrin adaptor CALM/PICALM is associated with neurofi brillary tangles and is cleaved in alzheimer’s brains. Acta Neuropathol 125(6):861-78. Ash PE, Bieniek KF, Gendron TF, Caulfi eld T, Lin WL, Dejesus-Hernandez M, van Blitterswijk MM, Jansen-West K, Paul JW,3rd, Rademakers R, et al. 2013. Unconventional translation of C9ORF72 GGGGCC expansion generates insoluble polypeptides specifi c to c9FTD/ALS. Neuron 77(4):639- 46. Asselta R, Rimoldi V, Siri C, Cilia R, Guella I, Tesei S, Solda G, Pezzoli G, Duga S, Goldwurm S. 2014. Glucocerebrosidase mutations in primary parkinsonism. Parkinsonism Relat Disord . Attems J, Jellinger KA, Lintner F. 2005. Alzheimer’s disease pathology infl uences severity and topographical distribution of cerebral amyloid angiopathy. Acta Neuropathol 110(3):222-31. Ayala YM, Pantano S, D’Ambrogio A, Buratti E, Brindisi A, Marchetti C, Romano M, Baralle FE. 2005. Human, drosophila, and C.elegans TDP43: Nucleic acid binding properties and splicing regulatory function. J Mol Biol 348(3):575-88. Bales KR, Du Y, Holtzman D, Cordell B, Paul SM. 2000. Neuroinfl ammation and alzheimer’s disease: Critical roles for cytokine/Abeta-induced glial activation, NF-kappaB, and apolipoprotein E. Neurobiol Aging 21(3):427,32; discussion 451-3. Barrett J, Fry B, Maller J, Daly MJ. 2005. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics 21(2):263-5. Beekes M, Th omzig A, Schulz-Schaeff er WJ, Burger R. 2014. Is there a risk of prion-like disease transmission by alzheimer- or parkinson-associated protein particles? Acta Neuropathol 128(4):463-76. Bertram L, McQueen MB, Mullin K, Blacker D, Tanzi RE. 2007. Systematic meta-analyses of alzheimer disease genetic association studies: Th e AlzGene database. Nat Genet 39(1):17-23. Bodmer WF. 1972. Evolutionary signifi cance of the HL-A system. Nature 237(5351):139,45 passim. Bogaerts V, Engelborghs S, Kumar-Singh S, Goossens D, Pickut B, van der Zee J, Sleegers K, Peeters K, Martin JJ, Del-Favero J, et al. 2007. A novel locus for dementia with lewy bodies: A clinically and genetically heterogeneous disorder. Brain 130(Pt 9):2277-91. Braak H and Braak E. 1991. Neuropathological stageing of alzheimer-related changes. Acta Neuropathol 82(4):239-59.

40 References

Brundin P, Melki R, Kopito R. 2010. Prion-like transmission of protein aggregates in neurodegenerative diseases. Nat Rev Mol Cell Biol 11(4):301-7. Bunton-Stasyshyn RK, Saccon RA, Fratta P, Fisher EM. 2014. SOD1 function and its implications for amyotrophic lateral sclerosis pathology: New and renascent themes. Neuroscientist . Cao X and Sudhof TC. 2001. A transcriptionally [correction of transcriptively] active complex of APP with Fe65 and histone acetyltransferase Tip60. Science 293(5527):115-20. Chapuis J, Hansmannel F, Gistelinck M, Mounier A, Van Cauwenberghe C, Kolen KV, Geller F, Sottejeau Y, Harold D, Dourlen P, et al. 2013. Increased expression of BIN1 mediates alzheimer genetic risk by modulating tau pathology. Mol Psychiatry 18(11):1225-34. Chen X, Hui L, Soliman ML, Geiger JD. 2014. Altered cholesterol intracellular traffi cking and the development of pathological hallmarks of sporadic AD. J Parkinsons Dis Alzheimers Dis 1(1):8. Chouraki V and Seshadri S. 2014. Genetics of alzheimer’s disease. Adv Genet 87:245-94. Clarimon J, Gray RR, Williams LN, Enoch MA, Robin RW, Albaugh B, Singleton A, Goldman D, Mulligan CJ. 2007. Linkage disequilibrium and association analysis of alpha-synuclein and alcohol and drug dependence in two american indian populations. Alcohol Clin Exp Res 31(4):546-54. Conibear E, Cleck JN, Stevens TH. 2003. Vps51p mediates the association of the GARP (Vps52/53/54) complex with the late golgi t-SNARE Tlg1p. Mol Biol Cell 14(4):1610-23. Cookson MR and van der Brug M. 2008. Cell systems and the toxic mechanism(s) of alpha-synuclein. Exp Neurol 209(1):5-11. Corder EH, Saunders AM, Risch NJ, Strittmatter WJ, Schmechel DE, Gaskell PC,Jr, Rimmler JB, Locke PA, Conneally PM, Schmader KE. 1994. Protective eff ect of apolipoprotein E type 2 allele for late onset alzheimer disease. Nat Genet 7(2):180-4. Cronin S, Hardiman O, Traynor BJ. 2007. Ethnic variation in the incidence of ALS: A systematic review. Neurology 68(13):1002-7. Cruchaga C, Karch CM, Jin SC, Benitez BA, Cai Y, Guerreiro R, Harari O, Norton J, Budde J, Bertelsen S, et al. 2014. Rare coding variants in the phospholipase D3 gene confer risk for alzheimer’s disease. Nature 505(7484):550-4. Cullen V, Sardi SP, Ng J, Xu YH, Sun Y, Tomlinson JJ, Kolodziej P, Kahn I, Saft ig P, Woulfe J, et al. 2011. Acid beta-glucosidase mutants linked to gaucher disease, parkinson disease, and lewy body dementia alter alpha-synuclein processing. Ann Neurol 69(6):940-53. Dawson TM and Dawson VL. 2003. Molecular pathways of neurodegeneration in parkinson’s disease. Science 302(5646):819-22. DeJesus-Hernandez M, Mackenzie IR, Boeve BF, Boxer AL, Baker M, Rutherford NJ, Nicholson AM, Finch NA, Flynn H, Adamson J, et al. 2011. Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron 72(2):245- 56. Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research, Saxena R, Voight BF, Lyssenko V, Burtt NP, de Bakker PI, Chen H, Roix JJ, Kathiresan S, Hirschhorn JN, et al. 2007. Genome-wide association analysis identifi es loci for type 2 diabetes and triglyceride levels. Science 316(5829):1331-6. Donnelly CJ, Zhang PW, Pham JT, Haeusler AR, Mistry NA, Vidensky S, Daley EL, Poth EM, Hoover B, Fines DM, et al. 2013. RNA toxicity from the ALS/FTD C9ORF72 expansion is mitigated by antisense intervention. Neuron 80(2):415-28.

41 References

Duyckaerts C and Dickson D. 2011. Neuropathology of alzheimer’s disease and its variants. In: Neurodegeneration: Th e molecular pathology of dementia and movement disorders. Dickson DW and Weller RO, editors. second ed. Wiley-Blackwell. 62 p. Population structure and aging [Internet]; c2014. Available from: http://epp.eurostat.ec.europa.eu/ statistics_explained/index.php/Population_structure_and_ageing . Feany MB and Bender WW. 2000. A drosophila model of parkinson’s disease. Nature 404(6776):394- 8. Felbecker A, Camu W, Valdmanis PN, Sperfeld AD, Waibel S, Steinbach P, Rouleau GA, Ludolph AC, Andersen PM. 2010. Four familial ALS pedigrees discordant for two SOD1 mutations: Are all SOD1 mutations pathogenic? J Neurol Neurosurg Psychiatry 81(5):572-7. Fillenbaum GG, van Belle G, Morris JC, Mohs RC, Mirra SS, Davis PC, Tariot PN, Silverman JM, Clark CM, Welsh-Bohmer KA, et al. 2008. CERAD (consortium to establish a registry for alzheimer’s disease) the fi rst 20 years. Alzheimers Dement 4(2):96-109. Fontenot AP, Torres M, Marshall WH, Newman LS, Kotzin BL. 2000. Beryllium presentation to CD4+ T cells underlies disease-susceptibility HLA-DP alleles in chronic beryllium disease. Proc Natl Acad Sci U S A 97(23):12717-22. Frost B and Diamond MI. 2010. Prion-like mechanisms in neurodegenerative diseases. Nat Rev Neurosci 11(3):155-9. Galpern WR and Lang AE. 2006. Interface between tauopathies and synucleinopathies: A tale of two proteins. Ann Neurol 59(3):449-58. Gibbs JR, van der Brug MP, Hernandez DG, Traynor BJ, Nalls MA, Lai SL, Arepalli S, Dillman A, Raff erty IP, Troncoso J, et al. 2010. Abundant quantitative trait loci exist for DNA methylation and gene expression in human brain. PLoS Genet 6(5):e1000952. Goate A, Chartier-Harlin M, Mullan M, Brown J, Crawford F, Fidani L, Giuff ra L, Haynes A, Irving N, James L. 1991. Segregation of a missense mutation in the amyloid precursor protein gene with familial alzheimer’s disease. Nature 349(6311):704-6. Goate A and Hardy J. 2012. Twenty years of alzheimer’s disease-causing mutations. J Neurochem 120 Suppl 1:3-8. Goedert M. 2001. Alpha-synuclein and neurodegenerative diseases. Nat Rev Neurosci 2(7):492-501. Goetz CG. 2000. Amyotrophic lateral sclerosis: Early contributions of jean-martin charcot. Muscle Nerve 23(3):336-43. Grabowski TJ, Cho HS, Vonsattel JP, Rebeck GW, Greenberg SM. 2001. Novel amyloid precursor protein mutation in an iowa family with dementia and severe cerebral amyloid angiopathy. Ann Neurol 49(6):697-705. Guerreiro RJ and Hardy J. 2011. Alzheimer’s disease genetics: Lessons to improve disease modelling. Biochem Soc Trans 39(4):910-6. Gundersen HJ, Bendtsen TF, Korbo L, Marcussen N, Moller A, Nielsen K, Nyengaard JR, Pakkenberg B, Sorensen FB, Vesterby A. 1988. Some new, simple and effi cient stereological methods and their use in pathological research and diagnosis. APMIS 96(5):379-94. Gusella JF, Wexler NS, Conneally PM, Naylor SL, Anderson MA, Tanzi RE, Watkins PC, Ottina K, Wallace MR, Sakaguchi AY. 1983. A polymorphic DNA marker genetically linked to huntington’s disease. Nature 306(5940):234-8.

42 References

Guyant-Marechal L, Rovelet-Lecrux A, Goumidi L, Cousin E, Hannequin D, Raux G, Penet C, Ricard S, Mace S, Amouyel P, et al. 2007. Variations in the APP gene promoter region and risk of alzheimer disease. Neurology 68(9):684-7. Haeusler AR, Donnelly CJ, Periz G, Simko EA, Shaw PG, Kim MS, Maragakis NJ, Troncoso JC, Pandey A, Sattler R, et al. 2014. C9orf72 nucleotide repeat structures initiate molecular cascades of disease. Nature 507(7491):195-200. Hamza TH, Zabetian CP, Tenesa A, Laederach A, Montimurro J, Yearout D, Kay DM, Doheny KF, Paschall J, Pugh E, et al. 2010. Common genetic variation in the HLA region is associated with late-onset sporadic parkinson’s disease. Nat Genet 42(9):781-5. Hardy J and Selkoe DJ. 2002. Th e amyloid hypothesis of alzheimer’s disease: Progress and problems on the road to therapeutics. Science 297(5580):353-6. Harold D, Abraham R, Hollingworth P, Sims R, Gerrish A, Hamshere ML, Pahwa JS, Moskvina V, Dowzell K, Williams A, et al. 2009. Genome-wide association study identifi es variants at CLU and PICALM associated with alzheimer’s disease. Nat Genet 41(10):1088-93. Hollingworth P, Harold D, Sims R, Gerrish A, Lambert JC, Carrasquillo MM, Abraham R, Hamshere ML, Pahwa JS, Moskvina V, et al. 2011. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with alzheimer’s disease. Nat Genet 43(5):429-35. Hood LE, Hunkapiller MW, Smith LM. 1987. Automated DNA sequencing and analysis of the human genome. Genomics 1(3):201-12. Hosler BA, Siddique T, Sapp PC, Sailor W, Huang MC, Hossain A, Daube JR, Nance M, Fan C, Kaplan J, et al. 2000. Linkage of familial amyotrophic lateral sclerosis with frontotemporal dementia to chromosome 9q21-q22. JAMA 284(13):1664-9. Hy LX and Keller DM. 2000. Prevalence of AD among whites: A summary by levels of severity. Neurology 55(2):198-204. Ince P. 2011. Dementia with lewy bodies and parkinson’s disease with dementia. Dickson DW and Wellers RO, editors. Second ed. Wiley-Blackwell. 224 p. Iqbal K, Alonso Adel C, Chen S, Chohan MO, El-Akkad E, Gong CX, Khatoon S, Li B, Liu F, Rahman A, et al. 2005. Tau pathology in alzheimer disease and other tauopathies. Biochim Biophys Acta 1739(2-3):198-210. Jonsson T, Atwal JK, Steinberg S, Snaedal J, Jonsson PV, Bjornsson S, Stefansson H, Sulem P, Gudbjartsson D, Maloney J, et al. 2012. A mutation in APP protects against alzheimer’s disease and age-related cognitive decline. Nature 488(7409):96-9. Jordan BD, Relkin NR, Ravdin LD, Jacobs AR, Bennett A, Gandy S. 1997. Apolipoprotein E epsilon4 associated with chronic traumatic brain injury in boxing. JAMA 278(2):136-40. Kachergus J, Mata IF, Hulihan M, Taylor JP, Lincoln S, Aasly J, Gibson JM, Ross OA, Lynch T, Wiley J, et al. 2005. Identifi cation of a novel LRRK2 mutation linked to autosomal dominant parkinsonism: Evidence of a common founder across european populations. Am J Hum Genet 76(4):672-80. Kalaria RN, Bhatti SU, Palatinsky EA, Pennington DH, Shelton ER, Chan HW, Perry G, Lust WD. 1993. Accumulation of the beta amyloid precursor protein at sites of ischemic injury in rat brain. Neuroreport 4(2):211-4. Kaneko H, Kakita A, Kasuga K, Nozaki H, Ishikawa A, Miyashita A, Kuwano R, Ito G, Iwatsubo T, Takahashi H, et al. 2007. Enhanced accumulation of phosphorylated alpha-synuclein and elevated beta-amyloid 42/40 ratio caused by expression of the presenilin-1 deltaT440 mutant associated with familial lewy body disease and variant alzheimer’s disease. J Neurosci 27(48):13092-7.

43 References

Kang J, Lemaire HG, Unterbeck A, Salbaum JM, Masters CL, Grzeschik KH, Multhaup G, Beyreuther K, Muller-Hill B. 1987. Th e precursor of alzheimer’s disease amyloid A4 protein resembles a cell- surface receptor. Nature 325(6106):733-6. Kara E, Lewis PA, Ling H, Proukakis C, Houlden H, Hardy J. 2013. Alpha-synuclein mutations cluster around a putative protein loop. Neurosci Lett 546:67-70. Kircher M, Witten DM, Jain P, O’Roak BJ, Cooper GM, Shendure J. 2014. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet 46(3):310-5. Kleppe K, Ohtsuka E, Kleppe R, Molineux I, Khorana HG. 1971. Studies on polynucleotides. XCVI. repair replications of short synthetic DNA’s as catalyzed by DNA polymerases. J Mol Biol 56(2):341-61. Knopman D. 2011. Clinical aspects of alzheimer’s disease. In: Neurodegeneration: Th e molecular pathology of dementia and movement disorders. Dickson DW and Weller RO, editors. second ed. UK: Wiley-Blackwell. 39 p. Kobayashi S, Tateno M, Park TW, Utsumi K, Sohma H, Ito YM, Kokai Y, Saito T. 2011. Apolipoprotein E4 frequencies in a japanese population with alzheimer’s disease and dementia with lewy bodies. PLoS One 6(4):e18569. Koistinaho M, Lin S, Wu X, Esterman M, Koger D, Hanson J, Higgs R, Liu F, Malkani S, Bales KR, et al. 2004. Apolipoprotein E promotes astrocyte colocalization and degradation of deposited amyloid-beta peptides. Nat Med 10(7):719-26. Kotzbauer PT, Cairns NJ, Campbell MC, Willis AW, Racette BA, Tabbal SD, Perlmutter JS. 2012. Pathologic accumulation of alpha-synuclein and abeta in parkinson disease patients with dementia. Arch Neurol 69(10):1326-31. Kruger R, Kuhn W, Muller T, Woitalla D, Graeber M, Kosel S, Przuntek H, Epplen JT, Schols L, Riess O. 1998. Ala30Pro mutation in the gene encoding alpha-synuclein in parkinson’s disease. Nat Genet 18(2):106-8. Kumar-Singh S, Cras P, Wang R, Kros JM, van Swieten J, Lubke U, Ceuterick C, Serneels S, Vennekens K, Timmermans JP, et al. 2002. Dense-core senile plaques in the fl emish variant of alzheimer’s disease are vasocentric. Am J Pathol 161(2):507-20. Kurland LT, Mulder DW, Westlund KB. 1955. Multiple sclerosis and amyotrophic lateral sclerosis (concluded); etiologic signifi cance of recent epidemiologic and genetic studies. N Engl J Med 252(17):697-702. Kwiatkowski TJ,Jr, Bosco DA, Leclerc AL, Tamrazian E, Vanderburg CR, Russ C, Davis A, Gilchrist J, Kasarskis EJ, Munsat T, et al. 2009. Mutations in the FUS/TLS gene on chromosome 16 cause familial amyotrophic lateral sclerosis. Science 323(5918):1205-8. Laaksovirta H, Peuralinna T, Schymick JC, Scholz SW, Lai SL, Myllykangas L, Sulkava R, Jansson L, Hernandez DG, Gibbs JR, et al. 2010. Chromosome 9p21 in amyotrophic lateral sclerosis in fi nland: A genome-wide association study. Lancet Neurol 9(10):978-85. Lamb H, Christie J, Singleton AB, Leake A, Perry RH, Ince PG, McKeith IG, Melton LM, Edwardson JA, Morris CM. 1998. Apolipoprotein E and alpha-1 antichymotrypsin polymorphism genotyping in alzheimer’s disease and in dementia with lewy bodies. distinctions between diseases. Neurology 50(2):388-91. Lambert JC, Ibrahim-Verbaas CA, Harold D, Naj AC, Sims R, Bellenguez C, DeStafano AL, Bis JC, Beecham GW, Grenier-Boley B, et al. 2013. Meta-analysis of 74,046 individuals identifi es 11 new susceptibility loci for alzheimer’s disease. Nat Genet 45(12):1452-8. Lambert JC, Heath S, Even G, Campion D, Sleegers K, Hiltunen M, Combarros O, Zelenika D, Bullido MJ, Tavernier B, et al. 2009. Genome-wide association study identifi es variants at CLU and CR1 associated with alzheimer’s disease. Nat Genet 41(10):1094-9. 44 References

Lander ES. 2011. Initial impact of the sequencing of the human genome. Nature 470(7333):187-97. Lautenschlager NT, Cupples LA, Rao VS, Auerbach SA, Becker R, Burke J, Chui H, Duara R, Foley EJ, Glatt SL, et al. 1996. Risk of dementia among relatives of alzheimer’s disease patients in the MIRAGE study: What is in store for the oldest old? Neurology 46(3):641-50. Lee HJ, Lee K, Im H. 2012. Alpha-synuclein modulates neurite outgrowth by interacting with SPTBN1. Biochem Biophys Res Commun 424(3):497-502. Lennox GG and Lowe JS. 1997. Dementia with lewy bodies. Baillieres Clin Neurol 6(1):147-66. Levy-Lahad E, Wijsman EM, Nemens E, Anderson L, Goddard KA, Weber JL, Bird TD, Schellenberg GD. 1995. A familial alzheimer’s disease locus on chromosome 1. Science 269(5226):970-3. Lin MK and Farrer MJ. 2014. Genetics and genomics of parkinson’s disease. Genome Med 6(6):48. Logroscino G, Traynor BJ, Hardiman O, Chio A, Mitchell D, Swingler RJ, Millul A, Benn E, Beghi E, EURALS. 2010. Incidence of amyotrophic lateral sclerosis in europe. J Neurol Neurosurg Psychiatry 81(4):385-90. Lv H, Jia L, Jia J. 2008. Promoter polymorphisms which modulate APP expression may increase susceptibility to alzheimer’s disease. Neurobiol Aging 29(2):194-202. Majounie E, Renton AE, Mok K, Dopper EG, Waite A, Rollinson S, Chio A, Restagno G, Nicolaou N, Simon-Sanchez J, et al. 2012. Frequency of the C9orf72 hexanucleotide repeat expansion in patients with amyotrophic lateral sclerosis and frontotemporal dementia: A cross-sectional study. Lancet Neurol 11(4):323-30. Marangi G and Traynor BJ. 2014. Genetic causes of amyotrophic lateral sclerosis: New genetic analysis methodologies entailing new opportunities and challenges. Brain Res . Mata IF, Samii A, Schneer SH, Roberts JW, Griffi th A, Leis BC, Schellenberg GD, Sidransky E, Bird TD, Leverenz JB, et al. 2008. Glucocerebrosidase gene mutations: A risk factor for lewy body disorders. Arch Neurol 65(3):379-82. McKeith IG, Dickson DW, Lowe J, Emre M, O’Brien JT, Feldman H, Cummings J, Duda JE, Lippa C, Perry EK, et al. 2005. Diagnosis and management of dementia with lewy bodies: Th ird report of the DLB consortium. Neurology 65(12):1863-72. Meeus B, Verstraeten A, Crosiers D, Engelborghs S, Van den Broeck M, Mattheijssens M, Peeters K, Corsmit E, Elinck E, Pickut B, et al. 2012. DLB and PDD: A role for mutations in dementia and parkinson disease genes? Neurobiol Aging 33(3):629.e5,629.e18. Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, Brownlee LM, Vogel FS, Hughes JP, van Belle G, Berg L. 1991. Th e consortium to establish a registry for alzheimer’s disease (CERAD). part II. standardization of the neuropathologic assessment of alzheimer’s disease. Neurology 41(4):479- 86. Mitsuyama Y. 1993. Presenile dementia with motor neuron disease. Dementia 4(3-4):137-42. Mok K, Traynor BJ, Schymick J, Tienari PJ, Laaksovirta H, Peuralinna T, Myllykangas L, Chio A, Shatunov A, Boeve BF, et al. 2012. Chromosome 9 ALS and FTD locus is probably derived from a single founder. Neurobiol Aging 33(1):209.e3,209.e8. Morgan K. 2011. Th e three new pathways leading to alzheimer’s disease. Neuropathol Appl Neurobiol 37(4):353-7. Mori K, Weng SM, Arzberger T, May S, Rentzsch K, Kremmer E, Schmid B, Kretzschmar HA, Cruts M, Van Broeckhoven C, et al. 2013. Th e C9orf72 GGGGCC repeat is translated into aggregating dipeptide-repeat proteins in FTLD/ALS. Science 339(6125):1335-8.

45 References

Myllykangas L, Polvikoski T, Reunanen K, Wavrant-De Vrieze F, Ellis C, Hernandez D, Sulkava R, Kontula K, Verkkoniemi A, Notkola IL, et al. 2002. ApoE epsilon3-haplotype modulates alzheimer beta-amyloid deposition in the brain. Am J Med Genet 114(3):288-91. Nalls MA, Pankratz N, Lill CM, Do CB, Hernandez DG, Saad M, DeStefano AL, Kara E, Bras J, Sharma M, et al. 2014. Large-scale meta-analysis of genome-wide association data identifi es six new risk loci for parkinson’s disease. Nat Genet 46(9):989-93. Neumann M, Kwong LK, Sampathu DM, Trojanowski JQ, Lee VM. 2007. TDP-43 proteinopathy in frontotemporal lobar degeneration and amyotrophic lateral sclerosis: Protein misfolding diseases without amyloidosis. Arch Neurol 64(10):1388-94. Newell KL, Hyman BT, Growdon JH, Hedley-Whyte ET. 1999. Application of the national institute on aging (NIA)-reagan institute criteria for the neuropathological diagnosis of alzheimer disease. J Neuropathol Exp Neurol 58(11):1147-55. Ng AS, Rademakers R, Miller BL. 2014. Frontotemporal dementia: A bridge between dementia and neuromuscular disease. Ann N Y Acad Sci . Nicoll JA, Roberts GW, Graham DI. 1995. Apolipoprotein E epsilon 4 allele is associated with deposition of amyloid beta-protein following head injury. Nat Med 1(2):135-7. Nicoll JA, Savva GM, Stewart J, Matthews FE, Brayne C, Ince P. 2011. Association between APOE genotype, neuropathology and dementia in the older population of england and wales. Neuropathol Appl Neurobiol 37(3):285-94. Noe E, Marder K, Bell KL, Jacobs DM, Manly JJ, Stern Y. 2004. Comparison of dementia with lewy bodies to alzheimer’s disease and parkinson’s disease with dementia. Mov Disord 19(1):60-7. Nowotny P, Simcock X, Bertelsen S, Hinrichs AL, Kauwe JS, Mayo K, Smemo S, Morris JC, Goate A. 2007. Association studies testing for risk for late-onset alzheimer’s disease with common variants in the beta-amyloid precursor protein (APP). Am J Med Genet B Neuropsychiatr Genet 144B(4):469-74. Ohtake H, Limprasert P, Fan Y, Onodera O, Kakita A, Takahashi H, Bonner LT, Tsuang DW, Murray IV, Lee VM, et al. 2004. Beta-synuclein gene alterations in dementia with lewy bodies. Neurology 63(5):805-11. Oinas M, Polvikoski T, Sulkava R, Myllykangas L, Juva K, Notkola IL, Rastas S, Niinisto L, Kalimo H, Paetau A. 2009. Neuropathologic fi ndings of dementia with lewy bodies (DLB) in a population- based vantaa 85+ study. J Alzheimers Dis 18(3):677-89. Pasanen P, Myllykangas L, Siitonen M, Raunio A, Kaakkola S, Lyytinen J, Tienari PJ, Poyhonen M, Paetau A. 2014. A novel alpha-synuclein mutation A53E associated with atypical multiple system atrophy and parkinson’s disease-type pathology. Neurobiol Aging 35(9):2180.e1,2180.e5. Peuralinna T, Oinas M, Polvikoski T, Paetau A, Sulkava R, Niinisto L, Kalimo H, Hernandez D, Hardy J, Singleton A, et al. 2008. Neurofi brillary tau pathology modulated by genetic variation of alpha- synuclein. Ann Neurol 64(3):348-52. Peuralinna T, Tanskanen M, Makela M, Polvikoski T, Paetau A, Kalimo H, Sulkava R, Hardy J, Lai SL, Arepalli S, et al. 2011. APOE and AbetaPP gene variation in cortical and cerebrovascular amyloid- beta pathology and alzheimer’s disease: A population-based analysis. J Alzheimers Dis 26(2):377- 85. Polvikoski T, Sulkava R, Haltia M, Kainulainen K, Vuorio A, Verkkoniemi A, Niinisto L, Halonen P, Kontula K. Apolipoprotein E, dementia, and cortical deposition of beta-amyloid protein. - N Engl J Med.1995 Nov 9;333(19):1242-7. . Polvikoski T, Sulkava R, Haltia M, Kainulainen K, Vuorio A, Verkkoniemi A, Niinisto L, Halonen P, Kontula K. 1995. Apolipoprotein E, dementia, and cortical deposition of beta-amyloid protein. N Engl J Med 333(19):1242-7. 46 References

Polvikoski T, Sulkava R, Myllykangas L, Notkola IL, Niinisto L, Verkkoniemi A, Kainulainen K, Kontula K, Perez-Tur J, Hardy J, et al. 2001. Prevalence of alzheimer’s disease in very elderly people: A prospective neuropathological study. Neurology 56(12):1690-6. Polymenidou M and Cleveland DW. 2011. Th e seeds of neurodegeneration: Prion-like spreading in ALS. Cell 147(3):498-508. Polymeropoulos MH, Lavedan C, Leroy E, Ide SE, Dehejia A, Dutra A, Pike B, Root H, Rubenstein J, Boyer R, et al. 1997. Mutation in the alpha-synuclein gene identifi ed in families with parkinson’s disease. Science 276(5321):2045-7. Purcell S, Neale B, Todd-Brown K, Th omas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, et al. 2007. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81(3):559-75. R Development Core Team. 2010. R: A language and environment for statistical computing and graphics [computer program]. Vienna, Austria: R Foundation for Statistical Computing. . Renton AE, Majounie E, Waite A, Simon-Sanchez J, Rollinson S, Gibbs JR, Schymick JC, Laaksovirta H, van Swieten JC, Myllykangas L, et al. 2011. A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron 72(2):257-68. Rioux JD, Xavier RJ, Taylor KD, Silverberg MS, Goyette P, Huett A, Green T, Kuballa P, Barmada MM, Datta LW, et al. 2007. Genome-wide association study identifi es new susceptibility loci for crohn disease and implicates autophagy in disease pathogenesis. Nat Genet 39(5):596-604. Rogaev EI, Sherrington R, Rogaeva EA, Levesque G, Ikeda M, Liang Y, Chi H, Lin C, Holman K, Tsuda T. 1995. Familial alzheimer’s disease in kindreds with missense mutations in a gene on chromosome 1 related to the alzheimer’s disease type 3 gene. Nature 376(6543):775-8. Rosen DR, Siddique T, Patterson D, Figlewicz DA, Sapp P, Hentati A, Donaldson D, Goto J, O’Regan JP, Deng HX. 1993. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362(6415):59-62. Ross OA, Toft M, Whittle AJ, Johnson JL, Papapetropoulos S, Mash DC, Litvan I, Gordon MF, Wszolek ZK, Farrer MJ, et al. 2006. Lrrk2 and lewy body disease. Ann Neurol 59(2):388-93. Sanger F, Nicklen S, Coulson AR. 1977. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74(12):5463-7. Schymick JC, Scholz SW, Fung HC, Britton A, Arepalli S, Gibbs JR, Lombardo F, Matarin M, Kasperaviciute D, Hernandez DG, et al. 2007. Genome-wide genotyping in amyotrophic lateral sclerosis and neurologically normal controls: First stage analysis and public release of data. Lancet Neurol 6(4):322-8. Serby M, Brickman AM, Haroutunian V, Purohit DP, Marin D, Lantz M, Mohs RC, Davis KL. 2003. Cognitive burden and excess lewy-body pathology in the lewy-body variant of alzheimer disease. Am J Geriatr Psychiatry 11(3):371-4. Sherrington R, Rogaev EI, Liang Y, Rogaeva EA, Levesque G, Ikeda M, Chi H, Lin C, Li G, Holman K, et al. 1995. Cloning of a gene bearing missense mutations in early-onset familial alzheimer’s disease. Nature 375(6534):754-60. Shlyueva D, Stampfel G, Stark A. 2014. Transcriptional enhancers: From properties to genome-wide predictions. Nat Rev Genet 15(4):272-86. Here’s How the ALS Ice Bucket Challenge Actually Started [Internet]. New York City: Time Inc.; c2014. Available from: http://time.com/3136507/als-ice-bucket-challenge-started/ . Sikorski AF, Sangerman J, Goodman SR, Critz SD. 2000. Spectrin (betaSpIIsigma1) is an essential component of synaptic transmission. Brain Res 852(1):161-6.

47 References

Sikorski AF, Terlecki G, Zagon IS, Goodman SR. 1991. Synapsin I-mediated interaction of brain spectrin with synaptic vesicles. J Cell Biol 114(2):313-8. Singleton AB, Wharton A, O’Brien KK, Walker MP, McKeith IG, Ballard CG, O’Brien J, Perry RH, Ince PG, Edwardson JA, et al. 2002. Clinical and neuropathological correlates of apolipoprotein E genotype in dementia with lewy bodies. Dement Geriatr Cogn Disord 14(4):167-75. Singleton AB, Farrer M, Johnson J, Singleton A, Hague S, Kachergus J, Hulihan M, Peuralinna T, Dutra A, Nussbaum R, et al. 2003. Alpha-synuclein locus triplication causes parkinson’s disease. Science 302(5646):841. Sisodia SS and St George-Hyslop PH. 2002. Gamma-secretase, notch, abeta and alzheimer’s disease: Where do the presenilins fi t in? Nat Rev Neurosci 3(4):281-90. Sleegers K, Brouwers N, Gijselinck I, Th euns J, Goossens D, Wauters J, Del-Favero J, Cruts M, van Duijn C, Van Broeckhoven C. 2006. APP duplication is suffi cient to cause early onset alzheimer’s dementia with cerebral amyloid angiopathy. Brain 129(11):2977-83. Soper JH, Kehm V, Burd CG, Bankaitis VA, Lee VM. 2011. Aggregation of alpha-synuclein in S. cerevisiae is associated with defects in endosomal traffi cking and phospholipid biosynthesis. J Mol Neurosci 43(3):391-405. Spillantini MG. 2011. Introduction to α-synucleinopathies. In: Neurodegeneration: Th e molecular pathology of dementia and movement disorders. Dickson DW and Weller RO, editors. Second ed. Blackwell Publishing Ltd. 191 p. Spillantini MG and Goedert M. 2000. Th e alpha-synucleinopathies: Parkinson’s disease, dementia with lewy bodies, and multiple system atrophy. Ann N Y Acad Sci 920:16-27. Spillantini MG, Crowther RA, Jakes R, Hasegawa M, Goedert M. 1998. Alpha-synuclein in fi lamentous inclusions of lewy bodies from parkinson’s disease and dementia with lewy bodies. Proc Natl Acad Sci U S A 95(11):6469-73. Sreedharan J and Brown RH,Jr. 2013. Amyotrophic lateral sclerosis: Problems and prospects. Ann Neurol 74(3):309-16. Strong MJ, Hortobagyi T, Okamoto K, Kato S. 2011. Amyothropic lateral sclerosis, primary lateral sclerosis and spinal muscular athropy. In: Neurodegeneration: Th e molecular pathology of dementia and movement disorders. Dickson DW and Weller RO, editors. Second ed. Blackwell Publishing Ltd. 418 p. Sun X, He G, Qing H, Zhou W, Dobie F, Cai F, Staufenbiel M, Huang LE, Song W. 2006. Hypoxia facilitates alzheimer’s disease pathogenesis by up-regulating BACE1 gene expression. Proc Natl Acad Sci U S A 103(49):18727-32. Tanskanen M, Makela M, Myllykangas L, Notkola IL, Polvikoski T, Sulkava R, Kalimo H, Paetau A. 2012. Prevalence and severity of cerebral amyloid angiopathy: A population-based study on very elderly fi nns (vantaa 85+). Neuropathol Appl Neurobiol 38(4):329-36. Tanskanen M, Lindsberg PJ, Tienari PJ, Polvikoski T, Sulkava R, Verkkoniemi A, Rastas S, Paetau A, Kiuru-Enari S. 2005. Cerebral amyloid angiopathy in a 95+ cohort: Complement activation and apolipoprotein E (ApoE) genotype. Neuropathol Appl Neurobiol 31(6):589-99. Tanskanen M, Kalaria RN, Notkola IL, Makela M, Polvikoski T, Myllykangas L, Sulkava R, Kalimo H, Paetau A, Scheltens P, et al. 2013. Relationships between white matter hyperintensities, cerebral amyloid angiopathy and dementia in a population-based sample of the oldest old. Curr Alzheimer Res 10(10):1090-7. Team RDC. R: A language and environment for statistical computing. Vienna, Austria:R Foundation for Statistical Computing. Available from http://www.R-project.org.

48 References

Trabzuni D, Ryten M, Walker R, Smith C, Imran S, Ramasamy A, Weale ME, Hardy J. 2011. Quality control parameters on a large dataset of regionally dissected human control brains for whole genome expression studies. J Neurochem 119(2):275-82. Trabzuni D, Wray S, Vandrovcova J, Ramasamy A, Walker R, Smith C, Luk C, Gibbs JR, Dillman A, Hernandez DG, et al. 2012. MAPT expression and splicing is diff erentially regulated by brain region: Relation to genotype and implication for tauopathies. Hum Mol Genet 21(18):4094-103. Tsuang D, Leverenz JB, Lopez OL, Hamilton RL, Bennett DA, Schneider JA, Buchman AS, Larson EB, Crane PK, Kaye JA, et al. 2013. APOE epsilon4 increases risk for dementia in pure synucleinopathies. JAMA Neurol 70(2):223-8. Tsubuki S, Takaki Y, Saido TC. 2003. Dutch, fl emish, italian, and arctic mutations of APP and resistance of abeta to physiologically relevant proteolytic degradation. Lancet 361(9373):1957-8. Van Deerlin VM, Leverenz JB, Bekris LM, Bird TD, Yuan W, Elman LB, Clay D, Wood EM, Chen- Plotkin AS, Martinez-Lage M, et al. 2008. TARDBP mutations in amyotrophic lateral sclerosis with TDP-43 neuropathology: A genetic and histopathological analysis. Lancet Neurol 7(5):409- 16. Vance C, Rogelj B, Hortobagyi T, De Vos KJ, Nishimura AL, Sreedharan J, Hu X, Smith B, Ruddy D, Wright P, et al. 2009. Mutations in FUS, an RNA processing protein, cause familial amyotrophic lateral sclerosis type 6. Science 323(5918):1208-11. Vehvilainen P, Koistinaho J, Gundars G. 2014. Mechanisms of mutant SOD1 induced mitochondrial toxicity in amyotrophic lateral sclerosis. Front Cell Neurosci 8:126. Vinters HV. 1987. Cerebral amyloid angiopathy. A critical review. Stroke 18(2):311-24. Viramo P and Sulkava R. 2010. Muistioireiden ja dementian epidemiologia. In: Muistisairaudet. Erkinjuntti T, Rinne J,Soininen H, editors. 1.th ed. Helsinki: Kustannus Oy Duodecim. 28 p. von Rotz RC, Kohli BM, Bosset J, Meier M, Suzuki T, Nitsch RM, Konietzko U. 2004. Th e APP intracellular domain forms nuclear multiprotein complexes and regulates the transcription of its own precursor. J Cell Sci 117(Pt 19):4435-48. Wang CS, Burke JR, Steff ens DC, Hulette CM, Breitner JC, Plassman BL. 2009. Twin pairs discordant for neuropathologically confi rmed lewy body dementia. J Neurol Neurosurg Psychiatry 80(5):562- 5. Wang W, Mutka AL, Zmrzljak UP, Rozman D, Tanila H, Gylling H, Remes AM, Huttunen HJ, Ikonen E. 2014. Amyloid precursor protein alpha- and beta-cleaved ectodomains exert opposing control of cholesterol homeostasis via SREBP2. FASEB J 28(2):849-60. Wider C, Ross OA, Nishioka K, Heckman MG, Vilarino-Guell C, Jasinska-Myga B, Erketin-Taner N, Rademakers R, Graff -Radford NR, Mash DC, et al. 2012. An evaluation of the impact of MAPT, SNCA and APOE on the burden of alzheimer’s and lewy body pathology. J Neurol Neurosurg Psychiatry 83(4):424-9. Wissemann WT, Hill-Burns EM, Zabetian CP, Factor SA, Patsopoulos N, Hoglund B, Holcomb C, Donahue RJ, Th omson G, Erlich H, et al. 2013. Association of parkinson disease with structural and regulatory variants in the HLA region. Am J Hum Genet 93(5):984-93. Wolfe MS, Xia W, Ostaszewski BL, Diehl TS, Kimberly WT, Selkoe DJ. 1999. Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and gamma-secretase activity. Nature 398(6727):513-7. Wu CA, Wang N, Zhao DH. 2013. An evaluation of the mechanism of ABCA7 on cellular lipid release in ABCA7-HEC293 cell. Chin Med J (Engl) 126(2):306-10.

49 References

Wuolikainen A, Andersen PM, Moritz T, Marklund SL, Antti H. 2012. ALS patients with mutations in the SOD1 gene have an unique metabolomic profi le in the cerebrospinal fl uid compared with ALS patients without mutations. Mol Genet Metab 105(3):472-8. Zaccai J, Ince P, Brayne C. 2006. Population-based neuropathological studies of dementia: Design, methods and areas of investigation--a systematic review. BMC Neurol 9(6):2. Zhang X, Zhou K, Wang R, Cui J, Lipton SA, Liao FF, Xu H, Zhang YW. 2007. Hypoxia-inducible factor 1alpha (HIF-1alpha)-mediated hypoxia increases BACE1 expression and beta-amyloid generation. J Biol Chem 282(15):10873-80. Zhao DH. 2007. Gap: Genetic analysis package. Journal of Statistical Soft ware 23(8).

50