Self-Assembly of Amyloid-ß Peptides in the Presence of Metal Ions and Interacting Molecules Self-Assembly of Amyloid- – a Detour of Amyloid Building Blocks

Cecilia Mörman Self-assembly of amyloid-ß peptides in the presence of metal ions and interacting molecules Self-assembly of amyloid- – a detour of amyloid building blocks

Cecilia Mörman ß peptides in ions and interacting presence of metal the molecules

Cecilia Mörman (former Wallin) received her MSc. in Biochemistry in 2015 from Stockholm University and continued with a PhD in the lab of Astrid Gräslund at Stockholm University. Cecilia’s research is mainly about amyloid-forming proteins.

ISBN 978-91-7911-188-5

Department of Biochemistry and Biophysics

Doctoral Thesis in Biophysics at Stockholm University, Sweden 2020

Self-assembly of amyloid-β peptides in the presence of metal ions and interacting molecules – a detour of amyloid building blocks Cecilia Mörman Academic dissertation for the Degree of Doctor of Philosophy in Biophysics at Stockholm University to be publicly defended on Thursday 3 September 2020 at 10.00 in Magnélisalen, Kemiska övningslaboratoriet, Svante Arrhenius väg 16 B.

Abstract Misfolding of proteins into amyloid structures is implicated as a pathological feature in several neurodegenerative diseases and the molecular causes are still unclear. One typical characteristic of Alzheimer’s disease is self-assembly and accumulation of soluble amyloid-β (Aβ) peptides into insoluble fibrils and plaques. One way to provide fundamental knowledge about the underlying fibrillization processes is to perturb the aggregation by varying the experimental conditions. Two main aspects are included in this thesis work: interactions with the Aβ peptide, and modulation of the Aβ peptide aggregation kinetics. The interplay between the Aβ peptide and three different types of aggregation modulators was studied mainly in vitro by biophysical techniques such as NMR, circular dichroism, and fluorescence spectroscopy. Metal ions, such as Ag(I), Cu(II), Hg(II), and Zn(II), at sub-stoichiometric concentrations with specific binding to monomeric Aβ peptides modulate and attenuate the Aβ self-assembly process. The bound (metal:Aβ) state removes Aβ monomers from the monomeric pool of amyloid building blocks used for fibril formation. In contrast, designed peptide constructs with cell-penetrating properties do not interact with monomeric Aβ, but exhibit an inhibitory effect on the Aβ oligomerization and fibrillization in vitro and in cells, via interactions with multimeric Aβ structures. The designed peptide constructs rescue Aβ-induced neurotoxicity and target both intracellular and extracellular Aβ. Full-length and native Tau protein, another protein implicated in Alzheimer’s disease, prevents the Aβ peptide fibrillization. The Aβ fibrillization process is not prevented by Tau interactions with the Aβ monomeric species, but rather with fibrils and oligomeric species of Aβ. Here we showed that the Aβ peptide interacts with various metal ions and molecules, both at the monomeric stage and as larger assemblies, with resulting perturbation of the Aβ aggregation kinetics. The interactions and aggregation modulators can be used to learn more about the underlying fibrillization processes and for the development of potential therapeutic strategies.

Keywords: biophysics, Alzheimer’s disease, protein aggregation, amyloid formation, amyloid-β peptide, aggregation kinetics, interactions, metal ions, designed peptide constructs, Tau protein, NMR, circular dichroism, fluorescence spectroscopy.

Stockholm 2020 http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-181495

ISBN 978-91-7911-188-5 ISBN 978-91-7911-189-2

Department of Biochemistry and Biophysics

Stockholm University, 106 91 Stockholm

SELF-ASSEMBLY OF AMYLOID-ß PEPTIDES IN THE PRESENCE OF METAL IONS AND INTERACTING MOLECULES

Cecilia Mörman

Self-assembly of amyloid-ß peptides in the presence of metal ions and interacting molecules

– a detour of amyloid building blocks

Cecilia Mörman ©Cecilia Mörman, Stockholm University 2020

ISBN print 978-91-7911-188-5 ISBN PDF 978-91-7911-189-2

Cover image: “11 arrows” by Cecilia Mörman. Aß amyloid aggregation is a heterogeneous process with several potential aggregation pathways, here symbolized by nine grey arrows (seven arrows are visible). These aggregation pathways may be disrupted to take a detour (pink arrows) by molecular interactions from aggregation modulators (pink). PDB ID 2MPZ, software’s PEP-FOLD3 and PyMOL.

Printed in Sweden by Universitetsservice US-AB, Stockholm 2020 ”Energy and persistence conquer all things” Benjamin Franklin

This thesis is dedicated to everyone fighting against Alzheimer’s disease

List of papers

Cecilia Mörman conducted most of the thesis work and articles under former name, Cecilia Wallin. This thesis is based on the following papers (I-VII):

I. Metal ion coordination delays amyloid-β peptide self-assembly by forming an aggregation-inert complex. Wallin C, Jarvet J, Biverstål H, Wärmländer S, Danielsson J, Gräslund A, Abelein A., 2020. Journal of Biological Chemistry, 295(21), pp.7224-7234. doi: 10.1074/jbc.RA120.012738. II. Characterization of Mn(II) ion binding to the amyloid-β peptide in Alzheimer’s disease. Wallin C, Kulkarni YS, Abelein A, Jarvet J, Liao Q, Strodel B, Olsson L, Luo J, Abrahams JP, Sholts SB, Roos PM, Kamerlin SC, Gräslund A, Wärmländer SK., 2016. Journal of Trace Elements in Medicine and Biology, 38, pp.183–193. doi: 10.1016/j.jtemb.2016.03.009. III. Mercury and Alzheimer’s Disease: Hg(II) Ions Display Specific Binding to the Amyloid-β Peptide and Hinder Its Fibrillization. Wallin C, Friedemann M, Sholts SB, Noormägi A, Svantesson T, Jarvet J, Roos PM, Palumaa P, Gräslund A, Wärmländer SKTS., 2019. Biomolecules, 10(1), pp.44. doi: 10.3390/biom10010044. IV. Specific Binding of Cu(II) Ions to Amyloid-Beta Peptides Bound to Aggregation-Inhibiting Molecules or SDS Micelles Creates Complexes that Generate Radical Oxygen Species. Tiiman A, Luo J, Wallin C, Olsson L, Lindgren J, Jarvet J, Roos P, Sholts SB, Rahimipour S, Abrahams JP, Karlström AE, Gräslund A, Wärmländer SK., 2016. Journal of Alzheimer’s Disease, 54(3), pp.971–982. doi: 10.3233/JAD-160427. V. Alzheimer’s disease and cigarette smoke components: effects of nicotine, PAHs, and Cd(II), Cr(III), Pb(II), Pb(IV) ions on amyloid- β peptide aggregation. Wallin C, Sholts SB, Österlund N, Luo J, Jarvet J, Roos PM, Ilag L, Gräslund A, Wärmländer SKTS., 2017. Scientific Reports, 7(1), pp.14423. doi: 10.1038/s41598-017-13759-5. VI. Designed Cell-Penetrating Peptide Inhibitors of Amyloid-beta Aggregation and Cytotoxicity. Henning-Knechtel A, Kumar S, Wallin C, Król S, Wärmländer SKTS, Jarvet J, Esposito G, Kirmizialtin S, Gräslund A, Hamilton AD, Magzoub M., 2020. Cell Reports Physical Science, 1(2), pp.100014. doi: 10.1016/j.xcrp.2020.100014. VII. The Neuronal Tau Protein Blocks in Vitro Fibrillation of the Amyloid-β (Aβ) Peptide at the Oligomeric Stage. Wallin C, Hiruma Y, Wärmländer SKTS, Huvent I, Jarvet J, Abrahams JP, Gräslund A, Lippens G, Luo J., 2018. Journal of the American Chemical Society, 140(26), pp.8138–8146. doi: 10.1021/jacs.7b13623.

The papers will be referred to as Paper I-VII in the thesis. All papers are reprinted with permission from the publishers.

List of additional papers – not included in the thesis

VIII. ATP impedes the inhibitory effect of Hsp90 on Aβ40 fibrillation. Wang H, Lallemang M, Hermann B, Wallin C, Loch R, Balzer B, Hugel T, Luo J. Submitted manuscript.

IX. Big dynorphin is a neuroprotector against amyloid-β peptide aggregation and cell toxicity. Gallego-Villarejo L, Suades-Sala A, Wallin C, Król S, Enrich Bengoa J, José Gomara M, Haro I, J. Muñoz F, Wärmlander S, Gräslund A, Perálvarez-Marín A. Submitted manuscript.

X. Studies on citrullinated LL-37: detection in human airways, antibacterial effects and biophysical properties. Al-Adwani S, Wallin C, Balhuizen MD, Veldhuizen EJA, Coorens M, Landreh M, Végvári Á, Smith ME, Qvarfordt I, Lindén A, Gräslund A, Agerberth B, Bergman P., 2020. Scientific Reports, 10(1), pp.2376. Original article.

XI. Metal binding to the amyloid-β peptides in the presence of biomembranes: potential mechanisms of cell toxicity. Wärmländer SKTS, Österlund N, Wallin C, Wu J, Luo J, Tiiman A, Jarvet J, Gräslund A., 2019. Journal of Biological Inorganic Chemistry, 24(8), pp.1189-1196. Mini-review.

XII. Pro-Inflammatory S100A9 Protein Aggregation Promoted by NCAM1 Peptide Constructs. Pansieri J, Ostojić L, Iashchishyn IA, Magzoub M, Wallin C, Wärmländer SKTS, Gräslund A, Nguyen Ngoc M, Smirnovas V, Svedružić Ž, Morozova-Roche LA., 2019. ACS Chemical Biology, 14(7), pp.1410-1417. Original article.

XIII. Copper ions induce dityrosine-linked dimers in human but not in murine islet amyloid polypeptide (IAPP/amylin). Dong X, Svantesson T, Sholts SB, Wallin C, Jarvet J, Gräslund A,

Wärmländer SKTS., 2019. Biochemical and Biophysical Research Communications, 510(4):520-524, pp.520–524. Original article.

XIV. Co-aggregation of pro-inflammatory S100A9 with α-synuclein in Parkinson’s disease: ex vivo and in vitro studies. Horvath I, Iashchishyn IA, Moskalenko RA, Wang C, Wärmländer SKTS, Wallin C, Gräslund A, Kovacs GG, Morozova-Roche LA., 2018. Journal of Neuroinflammation, 15(1), pp.172. Original article.

XV. Amyloid-β Peptide Interactions with Amphiphilic Surfactants: Electrostatic and Hydrophobic Effects. Österlund N, Kulkarni YS, Misiaszek AD, Wallin C, Krüger DM, Liao Q, Mashayekhy Rad F, Jarvet J, Strodel B, Wärmländer SKTS, Ilag LL, Kamerlin SCL, Gräslund A., 2018. ACS Chemical Neuroscience, 9(7), pp.1680–1692. Original article.

XVI. Mechanism of Peptide Binding and Cleavage by the Human Mitochondrial Peptidase Neurolysin. Teixeira PF, Masuyer G, Pinho CM, Branca RMM, Kmiec B, Wallin C, Wärmländer SKTS, Berntsson RP, Ankarcrona M, Gräslund A, Lehtiö J, Stenmark P, Glaser E., 2018. Journal of Molecular Biology, 430(3), pp.348–362. Original article.

XVII. The Amyloid-β Peptide in Amyloid Formation Processes: Interactions with Blood Proteins and Naturally Occurring Metal Ions. Wallin C, Luo J, Jarvet J, Wärmländer SKTS, Gräslund A., 2017. Israel Journal of Chemistry, 57(7), pp.674–685. Review.

XVIII. Ancient water bottle use and polycyclic aromatic hydrocarbon (PAH) exposure among California Indians: a prehistoric health risk assessment. Sholts SB, Smith K, Wallin C, Ahmed TM, Wärmländer SKTS., 2017. Environmental Health: A Global Access Science Source, 16(1), pp.61. Original article.

XIX. Self-Assembled Cyclic d,l-α-Peptides as Generic Conformational Inhibitors of the α-Synuclein Aggregation and Toxicity: In Vitro and Mechanistic Studies. Chemerovski-Glikman M, Rozentur-Shkop E, Richman M, Grupi A, Getler A, Cohen HY, Shaked H, Wallin C, Wärmländer SK, Haas E, Gräslund A, Chill JH, Rahimipour S., 2016. Chemistry - A European Journal, 22(40), pp.14236–14246. Original article. Abbreviations

Aβ Amyloid-β peptide LLPS Liquid-liquid phase separation AβPP Amyloid precursor protein LTP Long-term potentiation ACH Amyloid cascade hypothesis MAP Microtubule-associated protein AFM Atomic Force Microscopy MCI Mild cognitive impairment ALS Amyotrophic lateral sclerosis MRI Magnetic resonance imaging AMPs Antimicrobial peptides NCAM Neural cell adhesion molecule CAC Critical aggregation concentration NFT Neurofibrillary tangles CD Circular dichroism NMR Nuclear magnetic resonance CFs Curvilinear fibrils (protofibrils) PAH Polycyclic hydrocarbons CNS Central nervous system PET Positron emission tomography CPMG Carr-Purcell-Meiboom-Gill PFG Pulse-field gradient CPP Cell-penetrating peptide pFTAA Pentameric formyl thiophene acetic acid Cryo-EM Cryo electron microscopy PrP Prion protein CSF Cerebrospinal fluid PTM Post-translational modifications gOs Globular amyloid oligomers RF Rigid fibrils

GuHCl Guanidine hydrochloric acid RH Hydrodynamic radius DMSO Dimethyl sulfoxide ROS Reactive oxygen species DTNB 5,5-dithio-bis-(2-nitrobenzoic acid) SEC Size exclusion chromatography FCS Fluorescence correlation spectroscopy SOD Superoxide dismutase HFIP Hexafluoroisopropanol SPM Scanning Probe Microscopy HSQC Heteronuclear single quantum coherence TEM Transmission Electron Microscopy IAPP Islet amyloid polypeptide TTR Transthyretin IDP Intrinsically disordered protein ThS Thioflavin S iPSC Induced pluripotent stem cells ThT Thioflavin T

Contents

List of papers ...... i

List of additional papers – not included in the thesis ...... iii

Abbreviations ...... v

Contents ...... vii

1. Introduction ...... 1

2. Background ...... 2 2.1 Association and dissociation of biomolecules ...... 3 2.2 Misfolding protein diseases ...... 3 2.3 Neurodegeneration ...... 5 2.3.1 Alzheimer’s disease – general description...... 5 2.3.2 Alzheimer’s disease – biochemical and molecular description ...... 6 2.4 The Aβ peptide – Introduction to the system ...... 7 2.5 Amyloid and fibril formation ...... 9 2.5.1 The cross-β structure ...... 10 2.5.2 Aggregation versus amyloid fibrillization ...... 11 2.5.3 Amyloid formation from monomers via intermediate structures ...... 11 2.5.4 Are amyloid-forming proteins toxic? ...... 12 2.5.5 Amyloid fibrillization kinetics ...... 13 2.5.6 Modulation of protein aggregation ...... 17 2.6 What this thesis is about ...... 17 2.7 Aim ...... 18

3. Materials and Experimental techniques ...... 19 3.1 Sample preparation...... 19 3.1.1 Sample preparation for aggregation kinetics experiments ...... 20 3.2 Spectroscopy ...... 22 3.2.1 Circular dichroism ...... 22 3.2.2 Fluorescence...... 23 3.2.3 UV/Vis spectroscopy...... 26 3.2.4 Nuclear magnetic resonance ...... 26 3.3 Atomic Force Microscopy ...... 29

4. Metal ions as protein aggregation modulators ...... 30 4.1 Metal-binding properties of Aβ peptide (Paper I-V) ...... 32 4.1.1 Binding mode and aggregation ...... 32 4.1.2 Metal ion binding in membrane-mimetics ...... 34 4.2 Metal-Aβ complexes are unable for incorporation into fibril ends (Paper I) ...... 34 4.2.1 Ag(I) ions attenuate Aβ fibrillization by interfering with fibril-end elongation ...... 35 4.2.2 Fibril growth attenuation originates from Aβ peptides bound in metal complexes .... 35 4.2.3 Model: Metal ion perturbation of the fibrillization process ...... 37 4.3 Formation of reactive oxygen species by metal:Aβ complexes (Paper IV) ...... 38 4.4 Outlook ...... 38

5. Small molecules and peptide construct interactions with Aβ peptide assemblies . 40 5.1 Hydrophobic compounds found in cigarette smoke (Paper V) ...... 40 5.2 Designed peptide constructs (Paper VI) ...... 40

6. Two interacting proteins implicated in Alzheimer’s disease – Aβ and Tau ...... 45 6.1 Full-length native Tau protein prevents Aβ fibrillization (Paper VII) ...... 47 6.2 Outlook ...... 48

7. Concluding remarks and future perspectives ...... 49 7.1 Outlook ...... 50

8. Populärvetenskaplig sammanfattning ......

9. Acknowledgements ......

10. References ......

1. Introduction

“A trophy carries dust. Memories last forever.” Mary Lou Retton

“Everything should be made as simple as possible, but not simpler” was once proposed by Albert Einstein. Is that true for solving the riddle of Alzheimer’s disease – by finding all the chemical puzzle pieces underlying progression of disease? One strategy to find out the answer is by usage of molecular biology in combination with interdisciplinary collaboration with more complex models. Alzheimer’s disease is a relatively new disease, first described by Alois Alzheimer in 1907 (1). Followed by the “boom” of molecular biology development, about 30 years of molecular insights into the disease have been provided. The amyloid cascade hypothesis (ACH) with self- assembly/misfolding of amyloid-β (Aβ) peptides was first postulated in the 1990’s (2). Since then an enormous amount of intensive studies and clinical trials with potential drug candidates including passive immunization strategies have been executed – but so far, no strong results of disease modifying treatments have been developed (3–8). The need for therapeutic interventions and prevention strategies is high as well as development of effective biomarkers for early disease detection. Knowledge and molecular descriptions of the pathology in terms of amyloids and protein misfolding are still subject to notable improvements. This thesis is built on the foundation of the ACH, focusing on the Aβ peptide and its interaction partners using biophysical techniques. Prevention of Alzheimer’s disease is a complex problem – what does studies about a small peptide provide? It provides a small piece of a complicated problem, following the strategy of targeting a complicated problem simply by disentangle it in series of soluble and simple questions/solutions. Nevertheless, studies of the Aβ peptide are not simple, and the challenges motivate scientists all over the world. The small piece of a complex problem is here presented as the detour of amyloid building blocks, from the French word of détour, to perturb the building blocks of amyloids via molecular interactions for a temporary detained state before entering its way back to the amyloidogenic path. However, the physiological relevance/benefits for Alzheimer’s disease patients lie in future work. The anticipation is to provide insights into a 40 or 42 fragment peptide (one protein out of 100,000 proteins in a human cell) – one protein that is of high importance for underlying molecular processes in Alzheimer’s disease.

1 2. Background

“There are many hypotheses in science which are wrong. That’s perfectly all right; they’re the aperture to finding out what’s right. Science is a self- correcting process. To be accepted, new ideas must survive the most rigorous standards of evidence and scrutiny.” Carl Sagan

The definition of rest from physics is by simple means something static that does not change over time. In opposite, from the physiological perspective even at rest thousands of molecular processes are running constantly. It is not always the same ones, rather constant interchangeable processes, adapting to the needs of the current environment. A eukaryotic cell consists of about 70% of water, but nevertheless, the environment inside a cell is typically extremely crowded. All is tightly organized, controlled and regulated to keep homeostasis of a 80-400 mg/l concentrated soup of macrobiomolecules (9). One type of biomolecules within a cell is proteins and are the most abundant molecules other than water in biology in general (10). An average human male body consists of about 16% of proteins. Proteins are not only limited to “luxurious” proteins such as myosin and actin that build muscles, other proteins maintain essential functions such as being involved in signaling, adaptive systems, communication, energy factories, enzymes, stability and integrity, processing of food in the digestive tract, storage, clearance, membrane builders, defense systems, and so forth. Proteins are beautiful structures with intriguing properties. Structure lies within the order of the amino acids, and structure and function are tightly intertwined. Without a proper fold the function of the protein may be compromised. Cells are optimized to have a certain concentration of molecules that are needed at certain times, with fast adaptation to new stimulations/stress by external conditions. Once the systems responsible for keeping everything in order are not functioning properly, the protein concentration may be increased with undesirable protein accumulation. “No one succeeds alone” is also applicable to the protein world, several functional proteins consist of multimer structures, or requires cooperation in tandem for work performance. Many chemical processes rely on interactions between two or more biomolecules. In other words, it is not only limited to proper structure and fold/function, also interactions and dynamics are essential. A second layer of complexity comes with the thermodynamic characteristics and affinity of those interactions and the response of such an interaction. The dynamics of protein-protein or protein-molecule interactions in cellular systems are complex. How does it work and how is it controlled – to avoid unwanted association and dissociation?

2 2.1 Association and dissociation of biomolecules Association and dissociation are two fundamental processes underlying chemical reactions. Interactions and chemical processes are essential for life – as life is known today – where systems drive for equilibrium. On the other hand, in the scenario of total equilibrium where Gibbs free energy is zero and no work is performed – then life does not exist. Energy barriers, gradients, different milieus within defined compartments are giving rise to systems with dynamic equilibria, utilizing chemical processes striving for equilibria with equal rates of reactions forward and backwards. The equilibria may be shifted with increased concentration of products if the reactant concentrations are increased, according to Le Chatelier’s principle. One of the strongest interactions without covalent binding in biological systems is the one of biotin (vitamin B7/vitamin H) and avidin (a glycoprotein in egg white), an interaction also utilized in molecular biology methods. This -15 interaction is in the order of a KD of 1.3x10 M and the association rate is fast (11). The example of biotin-avidin is extraordinary, many biological complexes forms with an association rate of <103 M-1s-1 to >109 M-1s-1 (12) -4 -7 corresponding to a KD value of approximately 10 to10 M. In a complex system such as the interior of a living cell, proteins need to find the appropriate interacting partner among many others. Some interactions are weak and non-specific, while other interactions are stronger and specific. Specific, electrostatic- and hydrophobic interactions are all important. The cell has several safeguards to keep order among all possible interactions such as quality control systems, concentration-dependences, crowded environment, association and dissociation rates, turnover rates (degradation) and regulated expression/transcription/translation as well as an advanced system with feedback loops (13). In summary, association and dissociation of biomolecules are essential to perform work and for life itself. What happens when processes and systems fail to keep the balance of interactions and what are the consequences? If not regulated it could lead to uncontrolled accumulation, aggregation and protein misfolding (13). This will be discussed in more detail in the next section.

2.2 Misfolding protein diseases Many peptides and proteins share a common tendency to misfold from their native state and self-assembly towards formation of aberrant aggregates not applicable for proteolysis. This statement is true not only for proteins in ‘misfolding protein diseases’, but most proteins aggregate under conditions close to extreme boundaries. Initiation of misfolding – loss of native protein folds to incorrect folds – may originate from several causes such as inherited genetic mutations, post-translational modification (PTMs), thermal changes,

3 translational errors, and high protein concentration (14). Protein concentrations above the critical aggregation concentration (CAC) are vulnerable for aggregation, and the CAC is not a static measure but rather dependent on and influenced by environmental and experimental factors. The fibrillization process is basically a non-equilibrium process, as the final products are not reversible back to the initial state. More than 30 different aggregating and amyloidogenic proteins are associated with contribution to a broad range of human diseases. Notably, these proteins are also expressed under normal physiological conditions, indicating both normal and aberrant behavior. Some of the proteins are intrinsically disordered proteins (IDPs) (15), while others are folded in their native state. Several aggregating/misfolding proteins in this category of proteins are disease-specific but overlap between different diseases occur. Both IDPs and IDP-like proteins such as the Aβ peptide, Tau and α-synuclein implicated in Alzheimer’s and Parkinson’s disease, as well as natively folded proteins such as superoxide dismutase (SOD), lysozyme and transthyretin (TTR) may undergo the paths of misfolding. Insulin and islet amyloid polypeptide/amylin (IAPP) are two other protein examples related to diabetes mellitus. “The mad cow” disease, Creutzfeldt-Jakobs disease, and Kuru involve protein misfolding of the prion protein (PrP) – a protein that in the native state is partially folded with unfolded regions. Up to date there is only one disease-modifying drug available to treat TTR amyloidosis named Tafamidis with mechanisms and mode of action described (16). There are several well-written reviews about misfolding proteins (17–21). Folding and unfolding of proteins are complex processes, where the instructions of the fold lie within the amino acid sequence. Unfolded proteins are especially vulnerable to misfolding (22, 23). One fundamental question is what the “seed” of protein misfolding underlying pathology is, in other words the factors initiating directed misfolding. Misfolding of proteins into stable aggregates may follow different pathways involving on-pathway or off- pathway structures and intermediates. Up to date the exact mechanisms of misfolding and intermediate structures simultaneously present are not yet elucidated. Observations from several different proteins suggest an initial point of water-soluble monomeric structures, natively disordered or unfolded, and a final state with insoluble and elongated stable aggregates in equilibrium with a few percent monomeric species. The conversion from the initial state towards the final, stable state is subject to substantial studies, and the metastable intermediate structures have gained a large interest during the past years because of their potential importance for disease development. This “intermediate state” remains a black box with little information behind the structures and precise properties. The intermediate states are heterogenous and transient, which makes those structures a challenge to study. Recent studies attempted to stabilize low molecular intermediate states such as dimers, trimers, hexamers, heptamers, and octamers (24–27), by innovative

4 methodology, but the biological relevance, how they are formed, and how they proceed along the fibrillization pathway remain elusive. Protein misfolding is not only part of diseases, misfolded protein species may form without any deleterious consequences. The cell has several safety systems (28), one of them being chaperone proteins. One category of chaperones are ATP-dependent heat shock proteins, such as Hsp70 and Hsp90 (29), that are induced during cellular stress. Other examples are the BRICHOS domain (8, 30–32), clusterin (33, 34), and DNAJB6 (35), to name a few. As many things, these protective systems are not improving by advanced ages, and unnecessary stress in the upper age range should be avoided.

2.3 Neurodegeneration Protein misfolding is a phenomenon not only limited to certain tissues, several misfolding protein diseases, or proteinopathies, are systemic and target the nervous system. Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis (ALS), Huntington’s disease, and Creutzfeldt- Jakob disease are a few examples of central nervous system (CNS) degeneration (36). Neurodegenerative diseases involve aggregate and plaque formation within the CNS and are irreversible and often severe, present as a global contemporary problem both on the individual level and for the society. Although neurodegenerative disorders differ in symptoms and progression, some common denominators exist. They are all slowly progressive and neuronal death and cellular inclusions are common findings. Proteins causing amyloidosis are known to interact with different cellular constituents and molecules (37–41). Studying misfolding and aggregation kinetics of amyloidogenic proteins have gained an increased interest since the recognition of this process being part in Alzheimer’s disease and other neurodegenerative diseases.

2.3.1 Alzheimer’s disease – general description Few things scare as much as losing memory. Alzheimer’s disease is the most prevalent type of dementia characterized by neuronal atrophy, loss of synaptic activity and memory impairment with a global prevalence of nearly 24-40 million people with increasing numbers (42, 43). The disease is often mentioned as one disease, but late onset Alzheimer’s disease is rather a trajectory of a multifaceted disease with different symptoms and progression rates. Advanced age is one of the major risk factors (44), despite this, Alzheimer’s disease is not part of normal ageing (45, 46). About a few percentages of all cases are familial cases with known genetic defects with pathogenic mutations (47, 48), whereas the rest of the cases are considered sporadic cases. The APOε4 allele is more frequently present in both familial

5 and sporadic cases and it causes an increased risk of developing Alzheimer’s disease by 3-4 or >10 times for one or two copies, respectively (49, 50). On the opposite, one mutation recognized as the Icelandic mutation in the Aβ amyloid precursor protein (AβPP) is described as protective (51, 52). From the clinical perspective, mild cognitive impairment (MCI) and Alzheimer’s disease are often diagnosed by memory deficiency evaluation tests and laboratory tests. There are not yet any disease-modifying treatments available, only about five (53) symptomatic treatments such as cholinesterase inhibitors (54) and memantine (55). On the histopathological level, senile plaques and neurofibrillary tangles (NFTs) are two typical features of the disease. However, presence of senile plaques is not only limited to Alzheimer’s disease but also to other diseases as well as in asymptomatic people (56, 57). Instead a stronger correlation between NFTs has been noticed in sporadic cases of Alzheimer’s disease (58). Accumulation of Aβ usually occurs in grey matter, from the neocortex to the hippocampal area and further brain areas (59).

2.3.2 Alzheimer’s disease – biochemical and molecular description The Aβ peptide is the major constituent of senile plaques, whereas NFTs or paired helical filaments consist of aggregated and hyperphosphorylated Tau proteins. Other components of senile plaques beside Aβ peptides are other proteins, nucleic acids, lipids and metal ions (40). Part of the laboratory test diagnosis of Alzheimer’s disease involves increased total Tau and phosphorylated Tau protein and changed Aβ42/Aβ40 ratio in cerebrospinal fluids (CSF) (60, 61). The Aβ40 isoform is normally more abundant compared to the Aβ42 variant. In Alzheimer’s disease the Aβ42 concentration decreases compared to healthy individuals. At advanced stages the neuronal atrophy in brains of patients is visible by structural and functional magnetic resonance imaging (MRI), and senile plaques and NFTs are observed by radiopharmaceuticals in positron emission tomography (PET) imaging and by histologic markers (62). In recent years PET imaging has been shown to be an accurate but costly tool for diagnosing patients and for following disease progression (57, 63). Low Aβ42 levels is one typical marker for pre-clinical stages (60). Biomarkers in CSF and blood are also used (60, 64, 65), and new biomarkers are under constant search and development for diagnostic certainty, such as p-Tau181 (66, 67). A decrease of the acetylcholine levels in the brain is a common feature of the patient. Overall, the molecular and cellular mechanisms underlying Alzheimer’s disease are not yet known (68), but the ACH points at accumulation and aggregation of the Aβ peptide as a primary cause or as a secondary consequence (2, 20). Beyond the presence of Aβ peptides as the major building block of senile plaques, a link towards the precursor protein of Aβ with described inherently genetic mutations with associated increased risk for

6 Alzheimer’s disease are further supported by Down syndrome/Trisomi-21 (an extra copy of chromosome 21) with similar cognitive decline at young age as Alzheimer’s disease patients. Interestingly, like the old saying of “sleep helps the brain clean itself”, the CSF flow during sleep contributes to a decrease in local Aβ concentrations (69). In Alzheimer’s disease patients, biochemical changes occur already decade(s) before onset of symptoms related to disease pathology (70). Elucidating the molecular events preceding symptoms of the pathology is therefore of high importance to be able to target the disease on an early stage (71, 72). Apart from the ACH, several other hypotheses in the field of Alzheimer’s disease have been postulated (73–75). The Tau hypothesis (76–78) postulates hyperphosphorylation and aggregation of Tau proteins as contributions to the pathology. Inflammation and oxidative stress are two other major aspects of the disease, both at early stages and during progression. In some part also related is the metal hypothesis of Alzheimer’s disease (79, 80) originated from observed metal dyshomeostasis and senile plaques enriched of endogenous metal ions such as copper, iron, zinc and calcium (~mM concentration). The molecular interplay between the ACH and loss of function and/or gain of toxicity from aggregated Tau protein (81) and if the two proteins act in concert is still under debate (2, 76). Alzheimer’s disease is often described with an extracellular location of senile plaques and Tau protein NFTs with an intracellular location, however an equilibrium of intracellular and extracellular Aβ has been proposed and observed (82, 83), as well as propagation/spread of Tau protein (84, 85).

2.4 The Aβ peptide – Introduction to the system The Aβ peptide, derived from two consecutive enzymatic cleavages of a transmembrane glycoprotein named AβPP mainly located in the plasma membrane, has been exposed to intensive research since the first description in 1984 by Glenner and Wong (86). The Aβ peptide comes in several isoforms ranging from 37-43 residues long, with the 40 and 42 isoforms being the most abundant ones (87). The AβPP is constitutively and ubiquitously expressed in neurons especially, and the Aβ peptide is not the only fragment released by enzymatic cleavages. Potential pathways and resulting fragments are illustrated in Figure 1. The Aβ peptide is subject to proteolysis by degrading enzymes such as neprilysin and insulin-degrading enzyme (88). A large proportion of released Aβ peptides are secreted to the extracellular space, but intracellular uptake via endocytic vesicles occur. The amphipathic Aβ peptide is predominantly present as random coil (89) or polyproline II-helical conformation in water solution (90), determined by circular dichroism (CD) spectroscopy and NMR chemical shifts typical for a random coil structure. In membrane-mimicking environments the Aβ peptide adopts α-helical secondary structures (91).

Figure 1. Precursor protein of the Aβ peptide – AβPP. (A) Processing of AβPP by enzymatic cleavages by α- or β-secretase followed by γ-secretase yields the non-amyloidogenic or amyloidogenic pathways. The Aβ peptide is released from the AβPP protein via the amyloidogenic pathway. In (B) is the primary sequence of the Aβ peptide shown, together with a few fundamental properties.

The native functions of AβPP and the Aβ peptide are not fully understood but the AβPP gene family is evolutionary conserved in vertebrates (92). Knock-outs of AβPP and the AβPP-related proteins are reported as lethal (93). A possible role in neural growth and during development is suggested as well as cell adhesion functioning (88, 93–97). AβPP interacts with divalent metal ions such as copper and zinc (79, 98), and the expression levels are adopted to meet the available metal ion concentration (99–104). In addition, AβPP also takes part in cell adhesion and in synaptogenesis (93). The Aβ peptide is not only associated with pathology, at low concentration it is also present during healthy conditions. It has been proposed to be part of memory formation, communication between neurons, long-term potentiation (LTP) in

8 hippocampus, cell survival, and synaptic processes (94–96, 105). The neurotropic features of Aβ also includes metal chelating- and antioxidant properties (94). To conclude, both AβPP and Aβ are reported to be important for the neuronal function, but exactly why and how it is regulated remains to be clarified. The properties of the Aβ peptide contribute to a variety of possible interaction probabilities with different molecules (40, 99, 106). The strongest interaction measured is the one between two identical Aβ peptides (107). Two hydrophobic segments in the primary sequence of the Aβ peptides enable both interaction with biological membranes and inter- and intramolecular self- interactions. Under physiological conditions the Aβ peptide concentration is in the nanomolar range with a saturation concentration in the µmolar range (96, 108). The N-terminal part of the Aβ peptide is hydrophilic and has three histidine residues. Histidine residues are often involved in coordinating metal ions via the nitrogen in the imidazole ring. The N-terminal segment is a flexible part of the Aβ peptide, also shown as not being part of the cross-β structure within some fibril structures (109). Despite not being incorporated into the fibril core, the N-terminal part is suggested to be important for the amyloid aggregation formation (110). One aspect of importance is coordination of redox-active metal ions able to generate free radicals and oxidative stress (111).

2.5 Amyloid and fibril formation

Self-assembly and polymerization processes (polymer from the Greek word of polus, meaning “many, much”) are both biologically relevant, and only sometimes associated with disease. Formation of stable and homogenous structures as building blocks of the cytoskeleton by polymerization of microtubules and actin is necessary to keep the architecture and functioning of a cell. Other examples are biofilm formation in bacteria and biosynthesis of stabilizing molecules such as cellulose and lignin in cell walls viable for plant cells. The story of the term amyloid started already in 1842 by Matthias Schleiden who named starch for “amyloid” (from the latin word amulum), and in 1854 by Rudolph Virchow who studied cerebral corpora amylacea from brain tissues (112) and named the structures “amyloid” after the Latin word for starch. Later, it was established the structures in the brain tissues consisted mainly of protein material. Amyloid, amyloid-like and fibril formation via self-assembly processes are both functional and pathological (113–115). Storage of hormones (116), stress granules, biofilm formation (114), spider silk (115, 117), artificial scaffold proteins, and chorion proteins constituting fish and insect egg shells (113), are a few examples of beneficial aggregation properties. Amyloid formation is not

9 only selective for proteins in misfolding protein diseases, nearly “all” proteins despite lack of sequence homology can aggregate and form amyloid structures in vitro – it is only a question of experimental conditions. The steps from monomeric and soluble protein species into insoluble amyloids are not understood in molecular detail but the field is moving forward (17, 53, 118, 119). In terms of linear aggregation, the increase of insoluble amyloid materials may originate from two processes; increased concentration of aggregate number and/or increased concentration of fibril mass (119).

2.5.1 The cross-β structure The amyloid motif consists of an insoluble and stable cross-β structure (17) (Figure 2).The cross-β includes β-sheets structures aligned in parallel and packed as repeating units perpendicular to the fibril axis to an elongated fibril (120), and is characteristically recognized with a typical birefringence pattern in polarized light (apple green) once stained with Congo red (121). There are several protein fibrils with cross-β structures and filaments presented and visualized by different methodologies, such as fiber diffraction, X-ray crystallography, cryo electron microscopy (cryo-EM) and solid-state NMR (122–129).

Figure 2. Aβ amyloid and Tau paired helical filament structures. (A) Cross-β structure of

Aβ42 fibril from solid-state NMR studies, PDB ID 2NAO (124). (B) Amyloid Aβ40 fibril NMR structure from an Alzheimer’s disease patient, PDB ID 2M4J (128). (C) Cross-section of an in vitro formed Aβ fibril determined by cryo-EM, PDB ID 5AEF (123). (D) Aβ42 fibril from solid- state NMR studies with two hydrophobic cores of dimer structures, PDB ID 5KK3 (129). (E) Structure of paired helical Tau filaments determined by cryo-EM, PDB ID 5O3L (127).

10 Mature amyloid fibrils usually have a diameter in the range of 2-20 nm and lengths to several micrometer (121). Interestingly, all structures are not identical. Differences exist between different protein isoforms, mutations, variable experimental conditions, in vitro vs samples from patients, and differences between different patients (128). The polymorphic behavior with different inter-residue interactions may reflect the heterogeneous aggregation pattern of these proteins, easily perturbed by changes in the reaction volume. In general, Aβ fibril structures have a hydrophobic core stabilized by strong inter-strand backbone hydrogen bonds and intramolecular hydrogen bonds. Other interactions contribute to the amyloid fibril as well, such as side chain interactions, electrostatic interactions, van der Waals interactions, pi-stacking and hydrophobic interactions (121).

2.5.2 Aggregation versus amyloid fibrillization The self-assembly and amyloid formation processes are chemical reactions like crystallization of solutes, forming new aggregates by nucleation or fragmentation processes. The amyloid-state is often referred to as an additional phase with high thermodynamic stability, distinguishable from the crystalline and glass transition state. A sample of amyloid fibrils is easily described as gel-like. The reaction and the rate of the self-assembly reaction is dependent on solubility and concentration, and supersaturated conditions are usually used in in vitro kinetics studies. Co-precipitates, such as salts, detergents, membranes, etc. also affect the amyloid formation (130–133). It is important to distinguish aggregation from amyloid fibril formation. Amyloid fibril formation results in crystalline, stable amyloid structures, whereas aggregation in general can include fibrillization, precipitation and formation of amorphous aggregates. Proteins able to form amyloid structures both in vitro and in vivo can also under certain conditions, such as very rapid association, form amorphous aggregates, a process dependent on co- precipitants. Eventually amorphous aggregates may form amyloid fibrils over time. On the contrary, in the absence of any trace metal ions or buffer (ionic strength), amyloid formation is seldom present (134). Precipitation of proteins by salt additions is not surprising since the solubility of proteins is limited under certain ionic strengths, and this phenomenon is extensively used in practice to “salt out” proteins during i.e. protein purification. Liquid-liquid phase separation (LLPS) has obtained an increased research interest in terms of amyloid and neurodegeneration with interesting features both in vitro and in cells (135–137).

2.5.3 Amyloid formation from monomers via intermediate structures The aggregation and self-assembly of the Aβ peptide are usually described as a nucleation-dependent process with distinct amyloid on-pathway or off-

11 pathway intermediates. Whereas the characteristics of the structures at the start and the end of the amyloid fibrillization pathway are well described (89, 138–140), the intermediate states are poorly defined. The definition of the intermediate structures varies greatly in literature in terms of the characteristics (size, structure, formation) and names. Oligomers and protofibrils are the most common terms. Low-molecular weight oligomers and high-molecular weight oligomers are sub-groups often used, based on the molecular mass. Globular amyloid oligomers (gOs), highly curvilinear fibrils/protofibrils (CFs), and rigid fibrils (RF) are other commonly used nomenclature (141). Usage of different definitions and variable nomenclatures by different research groups is a limitation of the research field. In this thesis oligomers are referred to having a size of 2-100 monomers, whereas protofibrils are larger than oligomers but smaller than long (micrometer, diameter of 5-15 nm) fibrils. Despite the large interest of amyloid oligomeric structures and a few ones (stabilized in vitro) structurally described (24–27, 142), the transition process from oligomeric species into protofibrils and fibrils is scarcely well known. In one study oligomers with fibrillar structures (fragments of fibrils?) correlated with Alzheimer’s disease, whereas oligomers with non-fibrillar structures did not (143).

2.5.4 Are amyloid-forming proteins toxic? Mechanistic details of amyloid toxicity are still scarce but many attempts to link amyloid formation to cytotoxicity (41, 144) and to capture the toxic response of amyloids in vivo have extensively been tried. Transgenic mice for modeling Alzheimer’s disease are often not suitable/optimal to capture human pathology (59, 145–147). However, innovative model systems such as induced pluripotent stem cells (iPSCs) as 3D models derived from fibroblast donations from patients and controls differentiated to the cell type of interest with preserved age are under development (148–152). Early formed pre- fibrillar aggregates are suggested as the most toxic species (145, 153, 154), whereas fibril surfaces function as catalytic sites for new aggregates. In terms of cytotoxicity for any toxicant or toxin, the origin of the observed effects may be considered as gain-of-toxicity or loss-of-function. Both induced toxicity (155, 156) and loss of function have been hypothesized (157, 158). The mechanisms of toxicity may also be specific or non-specific. The physiological function of native Aβ peptides is not yet well known (section 2.4), but Aβ has been proposed as an endogenous antimicrobial peptide (AMP) in comparison with the human cathelicidin LL-37 (159–163). The AMP properties of the Aβ peptide is a shared feature with α-synuclein associated with Parkinson’s disease (164). Cytotoxicity in relation to the Aβ peptide is a complex question, especially when considering all secondary effects (165, 166). Several toxicity paths connected to Aβ aggregation and amyloid formation studied in several model systems have been proposed (17,

12 59, 145, 167, 168). Both interference and disruption of vital cellular functions and induced toxicity have been suggested. More precise suggestions involve mitochondrial dysfunction, interference with calcium homeostasis, perturbed cell membrane integrity, affected lipid and membrane composition, disruption of synaptic plasticity, inhibition of LTP, dysfunctional interactions with neuronal receptors and other proteins, excitotoxicity, pore formation, oxidative stress via reactive oxygen species (ROS) formation, inflammation, lipid peroxidation, and induced apoptosis (24, 25, 88, 146, 166, 169–178). The Aβ peptide toxicity is also dependent on several other factors according to a few studies, such as APOε4, cholesterol, mitochondrial network and clusterin (145, 179–186). The toxic origins of amyloid formation described in the previous section are under debate and widely discussed (154). Toxicity studies are complex, and the dose makes the poison as according to Paracelsus. Toxicity determination of potential toxic species should be used with caution, as different end-points of toxicity in different studies are often used, and overexpression and physiologically irrelevant (high) concentrations of the proposed toxicant to induce measurable end-points might not reflect pathological relevant events. On the other hand, detailed studies of antibodies in clinical trials showed an in vivo effect reflected in oligomeric species in vitro (187). Oligomeric structures with β-barrel pore-resembling structures have been described as well as larger assemblies with different properties (24–26, 115, 174). A large extent of the focus lies on oligomers with β-structures, whereas the β-content has not been correlated to toxicity (171) and early oligomeric structures have also been proposed to consist of α-structures (188). An interesting aspect discussed in a review of Eisele and Duyckaert is potential effects due to diffusion and spread of soluble oligomers versus insoluble oligomers, where the insoluble and diffusible oligomeric fraction may be the harmful one (59). Protein aggregation overload is another aspect of uncontrolled events with accompanying cell toxicity (81). Further, upstream effects of Tau protein by gingipains originated from the bacterium Porphyromonas gingivalis is another subject of interest in Alzheimer’s disease (189).

2.5.5 Amyloid fibrillization kinetics The fundamental processes underlying polymerization/fibrillization and filamentous growth are a combination of contributing processes (Figure 3) such as nucleation processes (primary and secondary), fragmentation, growth processes and dissociation processes (121, 190). Fragmentation of existing fibrils may be introduced to the system by shaking or stirring conditions, or simple from brittle fibrils breaking apart. Further, fragmentation is one of the main processes underlying growth of prion amyloid aggregates (191). Primary

13 nucleation describes a simple polymerization process, where nuclei of two or more monomers are formed with a higher association rate compared to the dissociation rate. In contrast, the dominating process underlying Aβ peptide fibrillization is related to monomer-dependent secondary nucleation process (192, 193). Secondary nucleation includes catalysis of new aggregates via the surface of existing fibrils or other surfaces (194). How the surface contributes to facilitated nucleation remains to be elucidated in full detail, but a role as a template or simply as a convenient meeting point for monomers has been suggested. The dissociation processes are usually described as negligible at the end of the fibrillization reaction, with an irreversible fibril state. The processes leading to an increase of fibril mass concentration during aggregation are mainly attributed to elongation processes (190). Growth of the Aβ fibril mass by addition of soluble and unstructured Aβ monomers to existing fibril ends is proposed by a dock-and-lock event, where soluble monomers binds to the fibril end and adopt to the conformation required for incorporation into the fibril in a two-step process (195–199). To obtain more information about the aggregation/fibrillization of proteins one strategy is to measure aggregate mass concentration as a function of time (119). The outcome and quality of such an experiment and resulting data is

Figure 3. Aggregation kinetics experiments. (A) Macroscopic, bulk experiments following a phenomenological sigmoidal behavior. Amyloid formation can be measured by a reporter dye recognizing amyloid structures. The lag phase is followed by the growth- and the saturation phase. During the lag phase the detection limit of small multimers and initial growth is not sensitive enough to capture those events. (B) Microscopic rate processes are distinguishably from the bulk experiments by global fit analysis with an integrated rate law (119, 192, 200).

14 highly dependent on the starting point/material, and it is crucial to control the starting material as careful and as accurate as possible. Different suitable techniques are available, and different methods (at least two) for validation of the results are recommended. The chosen technique should report linearly of the aggregate/fibril mass for proper investigation (201). Circular dichroism (CD) and amyloid-reporting fluorescent dyes (such as Thioflavin T (ThT) (202, 203) and pentameric formyl thiophene acetic acid (pFTAA) (204)) are two biophysical methods commonly used to measure amyloid formation as a function of time. ThT should not be confused with related Thioflavin S (ThS) molecules often used to stain biological materials for amyloid. Here I will focus on the usage of amyloid dyes for monitoring the fibrillization and to follow the aggregate mass concentration over time. A fast and simple amyloid kinetics analysis consists of phenomenological models such as sigmoidal curve fitting of bulk experiments with ThT/pFTAA fluorescence intensity measured over time to obtain phenomenological parameters. This type of analysis gives information of the evaluated kinetic curves for parameters that describe the curves as such, like aggregation halftime τ½, aggregation lag time, maximum growth rate, and end-point fluorescence intensity – but it gives no further information about the mechanisms behind the observations. A typical resulting sigmoidal kinetic trace shows a reasonable stable lag phase, but below the detection limit many molecular reactions are taking place (205). The lag phase is not only a “transportation phase”. Several molecular events are taking place not captured by typical bulk ThT/pFTAA fluorescence experiments. Single molecule techniques may be more successful. It was recently reported that about 60 monomers in an Aβ multimer was necessary for ThT-activity detected in fluorescence correlation spectroscopy (FCS) experiments (206). In recent years several mathematical models have been developed to describe the protein aggregation/fibrillization kinetics in mechanistic detail (119, 192, 193, 207– 209) facilitating usage of bulk experiments to obtain information about the microscopic rate processes of primary (kn) and secondary nucleation (k2), elongation (k+) and fragmentation (k-) (Figure 3 and 4). These models use rate laws to describe the chemical processes underlying fibrillization, and by taking all necessary events (reasonable number of events – still a simplified model of the observable) into account such as the concentrations of the reactants and products, an integrated rate law, or the Master equation, can be derived (118, 121, 193). Perturbation of the system by varying the experimental conditions such as concentration or by modulators gives a tool to study and elucidate the mechanisms behind the fibrillization process. Rome was not built in one day, likewise mechanistic information about protein fibrillization is not performed in one hand turn. With the limitations of a model, especially with the corresponding simplifications for a complex process such as protein aggregation, the quality and quantity of the data is extremely important, as well as a careful global fit analysis. On the other hand,

15 if successful, the global fit analysis allows determination of the microscopic rate constants giving rise to fibrillization as well as physical properties of the system (201).

Figure 4. Fit of microscopic rate constants for the processes of primary nucleation (kn), secondary nucleation (k2), and elongation (k+) from the influence of an aggregation inhibitor visualized as the nucleation dependence. The fit/modeling of the kinetic equations was performed in Amylofit online software (201).

Even more mechanistic information about the monomer-dependence and the reaction order of the fibrillization/nucleation process can be obtained using the scaling component and the mechanisms behind explaining the observables (119). The scaling component, γ, is obtained by a double logarithmic plot of τ½ versus the initial monomeric concentration, where a linear relationship reveals a slope equivalent to the γ-value. The scaling component is related to the reaction orders (201) and connects the macroscopic behavior observed in bulk experiments to the microscopic processes and helps the choice of suitable models for a deeper analysis of the data (121, 201). The Aβ peptide aggregation behavior in bulk fluorescence experiments differs depending on the length of the peptide. Compared to the Aβ40 peptide, Aβ42 is the more aggregation-prone one and only a tenth of the concentration is needed for fibrillization compared to the Aβ40 variant. On the microscopic level, for both peptide isoforms, the dominating processes underlying fibrillization are secondary nucleation processes. There is a large body of studies investigating the scaling component for the Aβ40 and Aβ42 peptides (190, 192) and it is slightly different. However, the differences (bulk) have also been described as reduced rate constants in general for Aβ40, and Aβ42 is not dependent on the monomer concentration for secondary nucleation processes to the same extent as Aβ40 (192). The dissociation rate of monomers from protofibrils, a slow process, have been reported to be very similar for the both Aβ peptide isoforms (210). Co-incubational studies with both Aβ40 and that Aβ42 have also been performed (61, 192, 211).

16 2.5.6 Modulation of protein aggregation Many factors affect the Aβ fibrillization processes. Variation of the experimental conditions such as pH, ionic strength, temperature, viscosity, molecular crowders, lipids, membrane mimetics, type of experimental tubes, small molecules, shaking conditions, other proteins, organic compounds, and Aβ local concentration may influence the aggregation. In a recent paper the influence of CSF on the Aβ aggregation kinetics was showed (212). The net negative charge of Aβ contributes to electrostatic repulsion between monomers which in turn make the peptides less aggregation-prone (213). Modulation of the fibrillization by organic or inorganic compounds (214, 215) such as curcumin (216), EGCG (217), metal ions (218), lipids (219), many organic compounds as metal chelators (220, 221), other FDA-approved drugs (222), β-sheet breakers (223), molecular tweezers, or Pt-based compounds (224, 225) has been extensively studied (39). Modulation of the Aβ amyloid aggregation kinetics by other proteins, such as chaperones, has also been studied (30, 31, 34, 35, 118, 226), as well as co-aggregation of two different amyloidogenic proteins (192, 227–232). The aggregation modulation can take place at different stages of the process, by induction of off-pathway structures or by interactions with monomers or with nuclei (kinetic intermediates) and fibrils. Interactions with monomers influence the available pool of aggregation-competent units. Coating of fibrillar surfaces by other proteins, such as the BRICHOS domain, efficiently attenuate secondary nucleation processes (31). Amyloid aggregation is partly driven by hydrophobic interactions. Small perturbations of the system may give rise to large effects in bulk experiments. Modulation of the aggregation kinetics both gives a tool for therapeutic strategies and a tool to gain more knowledge about the underlying fibrillization system by perturbing it. Noteworthy, assumptions based on whether the modulator concentration is decreased (consumed) over the aggregation time or not needs to be considered. Dual effects, such as observed changed aggregation kinetics in vitro and improved cognitive functions in clinical trials, should not be ignored.

2.6 What this thesis is about An increased understanding of the underlying processes of protein misfolding is central for medicine, material sciences and life sciences. This work is mainly built on pure biophysical strategies to understand Aβ aggregation and the interactions underlying the aggregation processes – how soluble peptide monomers aggregate into highly ordered structures and how this process can be modulated to take a detour from the fibrillization pathway by perturbational effects by metal ions and interacting molecules (Figure 5).

17 Copper-, zinc, and iron ion interactions with the Aβ peptide are not new to the field, rather the opposite – the Aβ peptide has been suggested to act as a weak metal chelator. Here we studied about 27 different metal ions to gain more information about the metal binding. In chapter 5 innovative peptide constructs were used to perturb the Aβ fibrillization, and the biophysical in vitro studies were moved into cell studies as well. In the last part, chapter 6, the interplay between two proteins implicated in Alzheimer’s disease (Aβ and Tau) was studied in vitro to provide new information about an interaction previously suggested from cellular and animal model studies. The choice of aggregation modulators was based on curiosity, the grade of modulating effects, and from ideas based on potential biological implications. We were both inspired by molecules with relationships to biological systems/diseases and of modulators providing information of amyloid aggregation per se.

Figure 5. Overview of the thesis. The thesis consists of two main topics, molecular interactions with monomeric Aβ peptides and modulation of the Aβ peptide aggregation behavior. These two topics are both separated and intertwined.

2.7 Aim

“It isn’t the mountains ahead to climb that wear you out; it’s the pebble in your shoe.” Muhammed Ali

Enormous efforts are put into research about misfolding protein diseases from several perspectives and interdisciplinary fields, but the biochemistry about these molecular processes are not yet elucidated. This thesis aims to contribute with further insights into the underlying Aβ peptide fibrillization processes probed by metal ions, designed peptide constructs and Tau protein. The overall aim is to understand the Aβ peptide’s molecular properties and interactions and how these interactions affect the Aβ’s aggregation behavior.

18 3. Materials and Experimental techniques

“Great things are not done by impulse, but by a series of small things brought together.” Vincent van Gogh

“In theory, there is no difference between theory and practice. But in practice, there is”. Manfred Eigen

The heart of a scientific paper is often referred to the methodology – a paper/conclusion is not better than the weakest link. In this section commonly used biophysical techniques to study conformational changes, interactions and other perturbations included in this thesis work are briefly described.

3.1 Sample preparation One important difference between ‘normal’ and amyloid aggregation prone proteins such as the Aβ peptide is that they instantly start to self-assemble during physiological conditions at concentrations frequently used for biophysical measurements. These features together with heterogenous mixtures of protein assemblies make it a challenge to study Aβ aggregation in a reproducible way, especially to understand the kinetics of amyloidogenesis and to characterize the kinetic intermediates. In order to study how Tau protein, designed peptide constructs, and metal ions affect the Aβ peptide and the aggregation kinetics the Aβ composition/aggregation state must be well distinguished in the experimental setup. A sample of Aβ peptides in buffer solution is a heterogeneous mixture with a stochastic nature of aggregation, especially at concentrations reaching the lower range for the critical CAC, and during the aggregation a distribution of many different structures eventually forming stable amyloid fibrils is present. In this section sample preparation is presented as protein sample preparation for experiments with already expressed and purified proteins. Optimization and good quality protein expression systems for amyloidogenic proteins are not part of this thesis, but in literature are several examples of such (233–235). One strategy for reproducible experiments is to take control over the initial state of the system as far as possible. This can be achieved by various approaches, such as additional protein purification to remove pre-formed aggregates (130). One approach is schematically presented in Figure 6. To begin with, buffers, co- precipitants and reagents need to be filtered before any use in experiments. In addition, unnecessary air-water interfaces and low-binding surfaces are avoided by using degassed buffers (130). Low-adhesive and metal-free tubes are recommended for the protein samples. Commercially available proteins are often offered both as recombinant proteins and as synthetic proteins. The cost may be higher for the recombinant proteins but are more accurate in terms of amino acid

19 composition and length. Important to keep in mind are the batch variation possibilities. The choice of buffer may vary depending on the experimental setup and the technique, further considerations involve suitable pH range, non- significant metal ion-buffer interactions, and compatibility with the method. Usage of a metal chelator for studies in the absence of metal ions is recommended due to the strong effect of metal ions on the aggregation kinetics. Notably, when metal ion interactions are the focus of the study, the metal ion solution may also be sensitive to buffer conditions and pH. Due to the aggregation propensity of Aβ peptides, lyophilized peptide powders are preferably dissolved at high pH (10 mM NaOH) or in an organic solvent such as Hexafluoroisopropanol (HFIP), Dimethyl sulfoxide (DMSO), or Guanidine hydrochloric acid (GuHCl). Organic solvents need to be removed before kinetic experiments, by gel filtration or evaporation, but low concentrations of DMSO (a few v/v %) may be used in kinetics experiments without any significant effects on the bulk experiment. Sonication of the protein in an ice-water bath for a couple of minutes (at least ~3 minutes) reduces the amount of pre-existing aggregates. Protein samples for experiments extra sensitive of the presence of pre-existing aggregates, such as aggregation kinetic experiments, need to be purified one step further.

3.1.1 Sample preparation for aggregation kinetics experiments In this section a generally used approach is briefly presented (Figure 6). The purpose of this step is to remove pre-formed aggregates to obtain good-quality samples for aggregation kinetics experiments:

- A high purity recombinant protein sample with a ~2 mg/ml concentration is dissolved in 6 M GuHCl and applied with a disposable syringe to a pre-equilibrated gel filtration column with a suitable running buffer for size exclusion chromatography (SEC) (i. e Superdex 75 10/300GL). - One column volume takes about 50 minutes (0.5 ml/min flow-rate), with sufficient separation for one peak with aggregated peptides and one peak corresponding to principally monomeric peptides (marked with a dashed circle in the chromatogram in Figure 6). The aliquots of purified proteins (filtrate) should directly be put on ice if the gel filtration is performed in room temperature. - The Aβ peptide has one Y10 residue, that may be used for protein concentration determination by absorbance spectroscopy with an extinction coefficient of 1424 M- 1cm-1. The apparent “free” protein concentration with several different equilibrium may affect the measured value. Once the concentration of the monomeric aliquots is measured, the sample can be used directly for kinetic experiments (preferable). Long storage times, i.e. to the next day or several days, may be done at +4 °C or -20 °C, respectively, but the quality of the sample may be compromised.

This step in the sample preparation for aggregation kinetics experiments takes time and with the cost of protein losses. However, this step is highly valuable in

Figure 6. Aβ sample preparation to avoid/reduce the number of pre-formed aggregates in monomeric Aβ samples for NMR and fibrillization kinetics experiments. The pre-formed aggregates and monomeric peptide visualized in the figure are derived from the protein data bank with PDB ID 2BEG, 2M4J, 5AEF, and 2M9S. The images are not to scale. The fluorescence correlation spectroscopy (FCS) results are subject for future publication (Mörman, Jarvet). order to be able to control the initial state before an aggregation kinetic experiments starts, for more reproducible and accurate measurements. A significant proportion of aggregated peptides in the sample is clearly removed by the SEC procedure. We measured the ThT-activity by FCS of samples obtained at different time points during the SEC (Figure 6). The running buffer did not contain any detectable ThT-active structures. A sample from the “aggregated

21 peak”, the peak prior the “monomeric peak”, did however show a high number of fluorescence intensity fluctuations (counts) corresponding to the presence of aggregates. In contrast, measurements of the “monomeric peak” did not show as many fluorescence intensity fluctuations, but interestingly still some counts indicative of a few structural assemblies with ThT-activity were present. Previous studies reported the smallest unit of Aβ multimers with ThT-activity as an aggregate of around 60 monomers (206). A CD spectrum of the “monomeric peak” shows typical random coil secondary structures, which is in line with most of the peptides being unstructured (Figure 6). However, the monomeric sample obtained after the SEC is in equilibrium with multimers with a few structures that are still ThT-active. A common tool to distinguish between primary and secondary nucleation processes (see section 2.5.5) is to add pre-formed seeds to the reaction volume with monomeric protein and different concentrations of an aggregation modulator at time zero. One way to generate seeds for seeding experiments include suspension of fibrils and sonication for a few minutes in an ice bath. Regarding the concentration of fibrils, one common way is to estimate the seed concentration based on the initial monomeric concentration. One important aspect regarding seeds is to avoid usage of old seeds. The concentration of seeds can be varied depending on the questions to be addressed. A low seeds concentration is useful to be able to conclude which one of the primary or secondary nucleation processes that is the dominating process, whereas a high seeds concentration (giving a concave shape instead of a sigmoidal curve) can be used for direct estimation of the elongation rate constant. The concentrations can vary between different experimental setups. For seeding experiments, it is important to use one stock solution of seeds that are distributed to all different conditions on the same plate for a fair comparison. Since the kinetics are faster in the presence of seeds, it is usually a good idea to measure data points more frequently. Different techniques to monitor aggregation kinetics are further presented in section 3.2.1 and 3.2.2.2.

3.2 Spectroscopy Spectroscopy is the study of interactions between electromagnetic radiation and matter. Here some of the methods used in the thesis are briefly described.

3.2.1 Circular dichroism The secondary and tertiary structure is an important feature of a protein, and for aggregation-prone amyloidogenic proteins with a transition from one conformation to another such information is crucial. Information about the secondary structure content is easily obtained from circular dichroism (CD) spectroscopy. CD is a fast and practical optical method. It is an ideal technique to

22 measure structural changes upon interactions, denaturing conditions, and changes in pH, temperature, salt concentration, and more. Many biomolecules are optically active. CD utilizes the property of chirality both regarding the molecule and to circularly polarized light. Chiral molecules interact and absorb left- and right circular polarized light to different extents. The ΔA=AL-AR difference spectrum consist of a pattern of characteristic structural features with both positive and negative bands possible. ΔA is often expressed as ellipticity, [θ]. For historical reasons θ=32.98ΔA. CD data is often presented as 2 mean residual molar ellipticity ([θ]MRME) with units of deg cm /dmol for a direct comparison of different samples regardless of concentration or size (236). For protein studies the far-UV region from ~180 to 260 nm is commonly used with the origin of the CD signal mainly by the peptide bond (237). The π-π* and n-π* transitions give typical bands at 190 and 210-220 nm, respectively (238). Typical CD spectra of secondary structures (random coil and β-structures) in proteins are shown in Figure 6. Spectral changes for longer wavelengths around 260-320 nm also occur for side chains of aromatic residues and disulphide bonds. Comparison with known structural spectra can easily be performed by using online software to calculate the percentage of i.e. α-helical content (239). The α-helical content is also easily derived by Eq. 1 (240).

훉 − 훉 훂 − 퐡퐞퐥퐢퐜퐚퐥 퐜퐨퐧퐭퐞퐧퐭 [%] = ( ퟐퟐퟐ 퐧퐦,퐫퐚퐧퐝퐨퐦 퐜퐨퐢퐥 ퟐퟐퟐ 퐧퐦,퐨퐛퐬퐞퐫퐯퐞퐝) ∗ ퟏퟎퟎ (Eq. 1) 훉ퟐퟐퟐ 퐧퐦,퐫퐚퐧퐝퐨퐦 퐜퐨퐢퐥 − 훉ퟐퟐퟐ 퐧퐦,훂−퐡퐞퐥퐢퐱

where (θ222 nm, random coil) is the average ellipticity for random coil structures in values of 3900 2 2 deg cm /dmol and the average ellipticity for α-helices is -35700 deg cm /dmol (θ222 nm, α-helix).

3.2.2 Fluorescence A common phenomenon used in a wide range of applications is fluorescence. Fluorescence is one type of luminescence, another example of luminescence is bioluminescence (sea-fire, luciferase in fireflies). Fluorescence is the name of the relaxation process back to the singlet ground state (S0) from an excited electronic state (often S1) via emission of a photon. The excited state is reached by absorption (excitation, 10-15 s) and internal conversion of a photon matching the energy between two energy levels. A fluorophore is a molecule/substance and chemical sensor with the property of absorbing light at a certain energy (wavelength) and emit a photon (fluorescence), often with a conjugated π-system. In protein chemistry, the aromatic residues tryptophan, tyrosine and phenylalanine are present as intrinsic fluorophores. Tryptophan and tyrosine are the ones mostly used for biochemical experiments. Extrinsic fluorophores may also be used, by labeling the protein of interest at a suitable place. Fluorescence measurements of a fluorophore is highly sensitive and specific, detecting only the fluorescence of interest, despite a potential large number of molecules absorbing light at the excitation wavelength. In addition, the fluorophore is sensitive to the local

23 environment, providing a tool for studying protein conformation, folding, misfolding, ligand binding, detection of localization etc. The quantum yield is the ratio of the number of emitted photons and the number of absorbed photons by the fluorophore, and it is temperature-dependent.

3.2.2.1 Interactions studies Relaxation as fluorescence is a spontaneous emission, competing with other processes (non-radiative). One of the competing processes is fluorescence quenching evolved from collision of the fluorophore, energy transfer, rearrangements, excited state or ground state reactions, or complex formation. The Aβ peptide has one tyrosine residue, Y10, and the intrinsic fluorescence may be used to study interactions with the Aβ peptide. Copper ions, Cu2+, are known fluorescence quenchers (241) and the Cu2+-dependent quenching can be used to study the interaction of the Aβ peptide with copper ions. The mechanism behind the fluorescence quenching of copper ions is not yet fully understood. The degree of Y10 fluorescence quenching upon titration with copper ions can be plotted against the [Cu], revealing a titration curve amenable for curve fitting with Eq. 2 for a 1:1 binding model (242) to determine the dissociation constant of the Cu:Aβ complex. In addition, studies of competitive binding with other metal ions without fluorescence quenching properties are also possible.

퐈 −퐈 ퟐ 퐈 = 퐈 + ∞ ퟎ ∙ (퐊퐚퐩퐩 + [퐂퐮] + [퐀훃] − √(퐊퐚퐩퐩 + [퐂퐮] + [퐀훃]) − ퟒ ∙ [퐂퐮] ∙ [퐀훃]) (Eq. 2) ퟎ ퟐ∙[퐀훃] 퐃 퐃

2+ where I∞ is the intensity upon saturation of Cu ions, I0 the initial intensity in the absence 2+ of Cu ions, and KD is the apparent dissociation constant.

3.2.2.2 Fibrillization studies The fluorescence phenomenon is utilized in amyloid aggregation kinetics experiments with extrinsic fluorophores recognizing amyloid structures. Commonly used dyes as probes for amyloid material are Congo red, Thioflavin S (ThS), Thioflavin T (ThT) (202, 203, 243, 244), and the luminescent conjugated oligothiophene pentameric formyl thiophene acetic acid (pFTAA) (204). ThT (λex 440 nm and λem around 480 nm) is a benzothiazole salt, a planar aromatic dye molecule that is highly flexible in solution where no fluorescence is detected due to relaxation processes of rotation. Once bound to amyloid material, the C-C bond is not as flexible anymore, and perhaps in combination with an increased hydrophobic environment, the fluorescence quantum yield increases to a high extent with emission at around 480 nm due to a bathochromic shift of the absorbance (245). Another probe, not as commonly used yet as ThT, is the pFTAA molecule (204) (λex 480 nm and λem around 520 nm) recognizing amyloid material, also with resulting emission fluorescence intensity changes from a restricted conformation upon binding to amyloid material (246). A comparison between ThT and pFTAA is shown in Figure 7. A similar binding site for pFTAA as for Congo red has been reported (246). In Paper I (supporting information,

24 Figure S2) the pFTAA influence on the Aβ aggregation kinetics was monitored, revealing no significant effect by the probe itself. However, another study reported changed fibril properties in the presence of higher concentrations of pFTAA (247).

Figure 7. Comparison of Thioflavin T (ThT) and pentameric formyl thiophene acetic acid

(pFTAA) in fibrillization kinetics experiments. 5 μM Aβ42 in 10 mM MOPS buffer pH 7.2 supplemented with 10 μM ThT or 0.3 μM pFTAA were used. The molecular structures for ThT (blue) and pFTAA (green) are shown in the right panel.

Both ThT and pFTAA may be used as probes to follow the amount of fibril mass formed over time. Such kinetics curves can be analyzed with phenomenological models, such as sigmoidal curve fitting, with Eq. 3 (248, 249).

퐀 퐅(퐭) = 퐅ퟎ + (Eq. 3) ퟏ+퐞퐱퐩 [퐫퐦퐚퐱(훕½−퐭)]

where F0 is the fluorescence baseline, A the amplitude of the kinetic curve, rmax is the growth -1 rate in h and the aggregation halftime is denoted τ½ in units of h. A more robust analysis of bulk experiments with parameters of biological meaning may be performed by a Master equation, an integrated rate law describing the evolution of fibril mass concentration by involved reaction rates (201, 207) (theoretical description of the approach is presented in section 2.5.5). This approach provides a tool to elucidate mechanistic information from aggregation kinetic curves. Back in time, Copernicus once said “Mathematics is written for mathematicians”, and luckily an online software with a user-friendly interface was developed by the lab of Knowles, Cambridge, UK (201). The Master equation is described as below, Eq. 4 (18, 118, 192, 193, 200, 207, 208, 250).

ퟐ 풌∞ 퐌(퐭) 푩 + 푪 푩 + 푪 풆휿풕 휿풌 + + − + ∞ −풌∞풕 = ퟏ − ( 휿풕 ) 풆 (Eq. 4) 퐌(∞) 푩+ + 푪+ 풆 푩− + 푪+

푛 +1 where 휅 = √2푘2푘+푚(0) 2 (or 푘2 = 푘− when n2=0) secondary nucleation 푛 휆 = √2푘푛푘+푚(0) 푐 primary nucleation (푘 ±푘̃ ) 휆2 퐵 = ∞ ∞ , 퐶 = ± , ± (2휅) ± (2휅2) 2 2 2 2 푘∞ = √2휅 /[푛2(푛2 + 1)] + 2휆 /푛푐 , 푘̃∞ = √푘∞ − 4 퐶+ 퐶− 휅

25 3.2.3 UV/Vis spectroscopy Many biomolecules do not absorb light in the visible region but absorb light in the UV region. This property is utilized for protein concentration determination, by aromatic residues (Trp, Tyr) around 280 nm, DNA at 260 nm and peptide bonds at around 214 nm. Some molecules, like ThT described in section 3.2.2.2 commonly used for detection of amyloidogenic material, they do absorb light in the visible region and this property is useful for accurate concentration determination. Formation of H2O2 was measured in Paper IV by a biochemical assay using 5,5-dithio-bis-(2-nitrobenzoic acid) (DTNB) and UV/Vis spectroscopy.

3.2.4 Nuclear magnetic resonance Nuclear magnetic resonance (NMR) spectroscopy is an excellent method to study biomolecules in liquid solution alike the natural environment with a broad range of applications such as for structural information, substance identification, inter- and intramolecular interactions, and one of the strengths of the method – to study dynamic processes (251) at equilibrium at several different timescales (Figure 8). Another strength of the method is the possibility to study weak intermolecular interactions. NMR provides high resolution information at the atomic level with the cost of low sensitivity due to a low energy difference between the ground state and the first excited state – connected to disadvantages of high amounts of materials needed and size limitation towards larger proteins, with low tumbling rate leading to line broadening and spectral overlaps. In NMR spectra, depending on the population/exchange rates, it might not be only one single conformation observed rather a weighted average from all observables. An atomic nucleus with a spin ≠ to zero has properties like a small magnet and can be perturbed by radiofrequency pulses in a strong, external magnetic field. This phenomenon is utilized in NMR spectroscopy, by perturbing the system in equilibrium towards an excited state and study the relaxation back to equilibrium. Relevant nuclei in protein samples with a spin quantum number of ½ includes 1H, 15N, 13C, and 31P. The natural abundances of 15N and 13C nuclei are low, requiring isotopically labeled samples for reasonable experimental times and detection. One further advantage with isotopically labeled proteins in terms of interaction studies is when the protein of interest can be labeled whereas the other protein/chemical is invisible in the spectrum to avoid signals from both. Atomic nuclei are sensitive to the chemical environment, and the chemical shifts and the amplitude of the signal resonances compared to a reference spectrum give information about structural changes, ligand-binding, exchange processes etc. As an example, a 1D or 2D NMR spectrum of a folded protein is rather dispersed, with nuclei with several different local and chemical environments, whereas an unfolded protein shows resonance signals with shifts very much alike. In other words, the chemical shifts of a protein provide a fingerprint spectrum of that particular protein under those experimental conditions. The resonance signals of the corresponding

26 residues can be assigned by measuring standard triple resonance backbone assignment experiments, such as HNCBCA and HN(co)CBCA. For the Aβ peptide there are several assignments already published (89, 91, 252, 253).

Figure 8. NMR timescales and molecular events. CPMG Carr-Purcell-Meiboom-Gill, NOE Nuclear Overhauser effect, PRE Paramagnetic relaxation enhancement, RDC Residual dipolar coupling.

Conformational changes and ligand-binding can be studied by NMR using chemical shift perturbation. As one example, a titration series of a ligand onto a 15N-labeled protein sample can be used and the chemical shift difference, Δδ, can be quantified by Eq. 5 (254, 255). The magnitude of the chemical shift differences Δδ does not always correspond with the strength of the effect (binding etc.).

½ ∆훅 ퟐ ∆훅 = ((( 퐍) + (∆훅 )ퟐ) /ퟐ) (Eq. 5) ퟓ 퐇

For an increased understanding of the equilibria of protein dynamics or ligand- binding kinetics in solution, information about the chemical exchange processes may be helpful (256). The exchange between two (or more) chemical states/environments are easily detected in the NMR spectrum by the rate relative to the frequency difference between an observable (257) such as the resulting chemical shift difference. The exchange rate (kex) between the states provides kinetic information, whereas the lifetime of such a state, the population (p), provides information about the thermodynamics of the dynamic process. Slow exchange (kex < Δω) relative to the NMR timescale with equally populated states stateA and stateB gives rise to two peaks (signal resonances) with the same intensity. With a faster, intermediate exchange rate (kex ~ Δω), the two states result in a broad peak in the middle of the former two peaks at slow exchange. In the fast exchange regime (kex > Δω), the exchange between the two states results in one narrow peak in the middle. Upon induced changes of nuclei’s chemical environment, such as ligand- binding, the signal resonances of the affected nuclei may lose signal intensity due to line broadening by paramagnetic effects or by chemical exchange. The loss of signal to “an invisible state” (258) can be further studied by relaxation dispersion experiments if the exchange occur on the ms-μs timescale, by gathering

27 information about the minor invisible state (stateB) from the kinetics of the properties of the major stateA. Compared to 2D heteronuclear single quantum coherence (HSQC) experiments where the monomeric major state, the unbound “free” state, is observed, Carr-Purcell-Meiboom-Gill (CPMG) relaxation dispersion experiments (259–261) provide information about the minor populated state, which can be the bound state when chemical exchange effects are present on the intermediate-fast NMR timescale (261–263). In some of my papers the bound state corresponds to the invisible state in 2D HSQC spectra, observed as loss of signal. The “invisible state”, the bound state, was studied by following the free, visible state by relaxation dispersion. In Paper 1 the chemical exchange between a metal:Aβ40 bound complex and 15 “free” Aβ40 was measured, assuming a two-state process. Pseudo 3D N-CPMG relaxation dispersion experiments were measured with a pulse scheme by Carr- Purcell-Meiboom-Gill with 11 different CPMG frequency delays between the refocusing 180° pulses. In Paper 1 one external magnetic field was used, but usage of two different external fields would have been optimal to determine the sign of Δδ. The transverse relaxation rates, R2, for each residue were calculated by Eq. 6.

퐨퐛퐬 ퟏ 퐈ퟎ 퐑ퟐ = ∙ 퐥퐧 ( ) (Eq. 6) 퐓퐂퐏 퐈

where TCP is the mixing time and I is the intensity of the peaks for each CPMG frequency delay with the reference spectrum TCP = 0 ms. CPMG-relaxation dispersion is observable in a narrow time-window.

In my studies with Aβ and metal ions, typically N-terminal residues gave rise to relaxation dispersion profiles. Affected residues do not automatically mean that they are directly involved (ligand) in the metal coordination, rather that the chemical environment around the nuclei are changed – perhaps due to metal coordination, or the outer shell of the coordination or structural changes. The relaxation dispersion data were further analysed with a two-state exchange model, thoroughly described in previous work by Axel Abelein (131, 264, 265) to obtain 0 information about the following parameters: pB, kex, Δδ, and 푅2. The concentration- and temperature dependence was also measured. Translational diffusion of a protein can be measured by NMR spectroscopy. By using the inhomogeneity in the applied magnetic field, by a magnetic field gradient pulse, it is possible to evaluate how nuclei diffuse in the tube and from peak intensity changes determine a diffusion coefficient. Pulse-field gradient (PFG) diffusion experiments regarding the Aβ peptide provides information to determine the translational diffusion coefficient, Dt, that can further be used to estimate the hydrodynamic radius (RH) via the Stokes-Einstein equation (266, 267) in Eq. 7. The approximation of RH here is determined as the radius of a spherical molecule.

28 퐤퐁퐓 퐃퐭 = (Eq.7) ퟔ훑훈퐑퐇

where kB is the Boltzmann constant, η is the viscosity, and T is the temperature.

3.3 Atomic Force Microscopy Microscopy is to see the invisible by the naked eye with magnification, both with visible light (optical microscopy), as well as with instrumental techniques utilizing other wavelengths (electron microscopy) or a probe to scan the surface of interest (scanning probe microscopy (SPM)). Atomic force microscopy (AFM) is one type of SPM. AFM utilizes the principle of measuring the force between a probe and the sample to gain information, such as 3D imaging topography, charge distribution, mechanical properties, electrochemical and magnetic properties, and spectral properties (IR/Raman). In the amyloid field AFM is commonly used to study the topographical properties of amyloid fibrils and aggregates/oligomers at high resolution (268), monomer-monomer interactions (269–271) and protein perturbation of lipid bilayers (272, 273). The measurements can be performed both in air and in liquid. One limitation of the AFM method is the sample preparation that includes drying of the samples onto the surface. Measurements in air on dried samples might not reflect the sample in solution. However, advances of the technique into “nanoscale” and even high-speed AFM providing a tool to directly study amyloid formation in real time (274) facilitate studies of amyloid formation. In this thesis AFM is mainly used to study the morphology of the amyloid end-products after aggregation kinetics measurements (Figure 9), for verification of fibril structures in the reaction volume. Such measurements can be performed on mica surfaces, a surface that is slightly negatively charged. In contact mode the surface is scanned at low speed with a constant setpoint/height/force mode. The image resolution and quality are dependent on the parameters set to adjust the feedback response. One disadvantage with contact mode is the risk of damaging either the tip or the surface. In tapping mode the tip on the cantilever vibrate (resonance frequency) and the cantilever is in a dynamic range between a contact and non-contact mode.

Figure 9. Topographical AFM image of Aβ40 fibrils recorded in tapping mode in air on a mica surface to visualize the fibril morphology. The scale bar corresponds to 1 μm, data to be published (Mörman, Luo).

29 4. Metal ions as protein aggregation modulators

“For me too, the periodic table was a passion. ... As a boy, I stood in front of the display for hours, thinking how wonderful it was that each of those metal foils and jars of gas had its own distinct personality.” Freeman Dyson

Metal ions – a large family of ions with both common and distinctive properties. 80% of the periodic table consists of metals and with several oxidation states the possible variation of metal ions results in many different ions. From a biological perspective, metal ions are essential and participate in biological reactions (275), or used for stability reasons necessary for life- sustaining processes. Specificity of a metal to a metal binding site is reached by thermodynamics and kinetics. Billions of years ago the first cells utilized metal ions abundantly present in their environment, mainly Fe2+ and Mg2+ (276). During evolution other metal ions became available also suitable for participating as co-factors in proteins, with an induced pressure for the cellular systems to adapt to the new milieu with several utilizable metal ions (276). About 50% of all proteins today utilize metal ions. In contrast, certain metal ion concentrations are kept under strict control in biological and cellular systems to avoid deficiency or toxic overload. Several d-block transition metal ions are present in the human brain. The “free” metal ions concentrations are very low intracellularly (for Zn2+ about 10-10-10-15 M and for Cu+ about 10-18 M (276)). But, increased transition metal ion levels (about 300 µM Zn ions and 15 µM Cu ions) (100, 277, 278) are released in the synaptic cleft during neuronal transmission – which is the same cellular compartment where the Aβ peptide is released from its precursor protein. Notably, compromised cellular synapse plasticity has been suggested to be involved in the Alzheimer’s disease neuronal loss (155, 166, 278). Metal dyshomeostasis is implicated in Alzheimer’s disease and in other misfolding protein diseases, with impaired biomolecular functioning, ROS formation, accumulation, elevated/changed concentrations, or mislocalization of metal ions (56, 79, 104, 166, 167, 279–290). The data in the literature are spread, sometimes with observed elevated/decreased metal ion concentrations (288, 291–303). Several challenges lie in the attempt to use complex model systems such as patients and healthy controls, with disperse disease progression, not optimal diagnosis opportunities, usage of non-age matched controls, and analytical methods and biomarkers with low sensitivity or high detection limits. Estimation of environmental exposure of both endogenous and contaminant metal ions is also a challenge, as the vulnerable time window (304) of acute or chronic exposure to harmful/toxic or deficient metal levels

30 may differ and might not be reflected by single measurements. There is growing evidence supporting binding of metal ions to amyloid species, specifically in the context of misfolding diseases (104, 218, 305–310) with or without a link to other factors (311). It is not yet completely elucidated if metal ion interactions with misfolding proteins in neurodegenerative diseases are a cause, a consequence or only secondary effects from disease progression/pathology (312). Correlation and causation are two important concepts to distinguish. One wonders if metal ion interactions with amyloidogenic proteins occur in healthy cells too, or only applies during pathological conditions. Chelation-based therapeutic strategies for neurodegenerative diseases such as Alzheimer’s disease have been tested and are also under development (221, 286, 308, 313–315). Metal ion interactions with the Aβ peptide (316) are indeed interesting and ions of particular interest have been copper, zinc, calcium and iron – metal ions that are abundantly sequestered in senile plaques (99, 299, 317, 318). Copper, manganese and iron are also highlighted metal ions due to their redox properties and possible induction of oxidative stress (111, 177, 242, 279, 319– 322). The Aβ peptide binds both Cu(II) and Cu(I) ions with different binding modes, and the transition between those two binding modes are of interest for an understanding of the molecular details of ROS formation (111, 177). The Aβ peptide interacts with certain metal ions in the nano- to µM range, binding metal ions at the N-terminus both as a monomer and as more structured aggregates (99, 106). Upon metal binding the net charge of the Aβ peptide is decreased which may facilitate self-assembly processes. With fast self-association amorphous aggregates can be induced instead of ordered amyloid structures. Discussions about metal ion inducible effects on the Aβ peptide is much dependent on the metal ion concentration versus the peptide concentration (323). When metalpeptide situation where non-specific interactions and electrostatic effects become more important. The metal- ligand binding is easily described as a competition between the metal ion and protons. The Aβ peptide lacks the typical metal binding cysteine residues (polarizable atoms of thiolate and carboxylate), but three histidine residues together with negatively charged residues in the N-terminus are typical metal binding ligands under physiological conditions in the human Aβ peptide variant. The histidine residues, as well as Asn and Gln, are polar, uncharged and favorable for metal binding (276). The histidine residues with a pKa of ~6 are however sensitive to changes of pH close to physiological pH (324, 325). In contrast, the murine Aβ peptide has only two histidine residues, together with Y10 substituted to a phenylalanine and R5 to a glycine. Surprisingly, the murine Aβ peptide display a higher affinity towards metal ions compared to the human one (326). The origin of this effect has been suggested to be the absence of a charged and somewhat bulky arginine residue. R5 in the wildtype

31 Aβ is not in the first coordination sphere, rather in the second one, but does still affect the metal ion interaction. Systematically studying metal ion interactions and modulation of the Aβ self-assembly can provide important information for the understanding of disease etiology (111, 312, 327). Pioneering work by Atwood and co-workers studied several different metal ions incubated with Aβ and consequently induced amounts of aggregates (328). However, information such as binding and detailed mechanisms were not provided to explain the observed effects. By studying the interaction in detail, pieces of information can possibly be put together to connect the binding to the metal-modulated aggregation behavior (307). Why certain metal ions bind to the Aβ peptide specifically and others do not is another question to be addressed. We studied about 27 different metal ions (Table 1) with molecular resolution to understand the biophysical effects on the aggregation behavior, where a pattern started to develop. Here we will discuss only a selection of these 27 metal ions (Paper I-V): Ag(I), Cd(II), Cr(III), Cu(II), Hg(II), Mn(II), Pb(II), Pb(IV), and Zn(II) ions. In addition to metal binding per se, metal ions are convenient to work with to perturb the Aβ peptide aggregation kinetics which allows further mechanistic insights into the aggregation and nucleation mechanisms. To study the metal-induced effects on the Aβ aggregation behavior, Ag(I) and Zn(II) were chosen as model ions and discussed in more detail in section 4.2.

4.1 Metal-binding properties of Aβ peptide (Paper I-V)

4.1.1 Binding mode and aggregation In Paper I-V a selection of metal ions was studied in terms of binding specificity. The metal ion:Aβ40 binding was studied with high-resolution NMR spectroscopy and molecular modeling simulations. One question addressed was why certain metal ions bind to the Aβ peptide specifically, while others do not. d-block transition metal ions are prominent protein binders via coordination or electrostatic interactions (329). Metal specificity is often thermodynamically controlled via ionic radii, coordination number, Lewis acid type, and coordination geometry. Polarizability of the ligand and metal also contributes (275). The Aβ peptide is intrinsically disordered and hence flexible to adjust its structure for coordination of different metal ions despite differences in sizes and charges. The binding mode is similar but not exclusive between the metal ions, rather slightly different and the second sphere coordination affects the specificity and hence the binding mode. Following a similar concept, there is not only one binding mode per metal ion; instead a collection of different but related populated binding modes with a

32 Table 1. Metal-binding properties of the Aβ peptide with fundamental features of 33 metal ions – 27 ions here experimentally studied in the thesis. The affinity values (dissociation constant, KD) for the metal:Aβ complexes vary depending on the experimental setup and the state of the sample. Here are the KD values estimated based on similar experimental conditions. Metal ions marked in green display specific binding towards Aβ. (330) (331) (332) (111, 333) (322)(334, 335) (40) (334)

* Yes for high metal ion concentrations (super stoichiometric), ** 1:1 ratio slower aggregation kinetics transient character are present. The imidazole group of histidine residues is on the borderline (intermediate) between being a “hard” or a “soft” electron donor, or base, according to the Pearson acid base concept (332, 336). From the data in Table 1 the Aβ peptide is suggested to be on the borderline towards a “soft” electron donor. None of the studied metal ions considered as “hard” acids (Cr(III), K(I), Na(I), Ca(II) (337), Mg(II), Al(III)) interact specifically with the Aβ peptide, but three different ions (Cu(I) (333), Ag(I), and Hg(II)) regarded as “soft” acids do (Paper I and III). Metal ions regarded as intermediate acids that bind specifically to the Aβ peptide are Cu(II), Zn(II), Ni(II), Co(II), Mn(II), and Fe(II) (322) (Table 1). One exception seems to be Pb(II) (Paper V), as it is a “borderline/intermediate” acid but does not bind specifically to the Aβ peptide. Based on these observations, the Aβ peptide can interact with a variety of metal ions that lies in the intermediate/soft acid region with different affinities. No binding towards trivalent ions were observed, only nonspecific electrostatic interactions. Two monovalent, one tetravalent, and nine divalent ions display specific binding towards the Aβ

33 peptide (Table 1), and eight of those ions studied here (with a moderate binding affinity) modulate the Aβ fibrillization kinetics. Cu(II) ions bind, at least to our current knowledge, definitely with the highest affinity towards the Aβ peptide, following the Irving-Williams series which is based on high-spin divalent ions and their corresponding ionic radii (ionic potential) and compares the stability of the metal complexes (and the Jahn-Teller effect). In biology, the intracellular concentrations of transition metal ions also follow the pattern of the Irving-Williams series. The metal ions with highest affinity towards bioligands are also kept at lowest concentrations via cellular regulation of the concentrations – to avoid unwanted competition between more weakly binding metal ions (276). The general assumption of the Aβ peptide being a typical divalent metal ion binder has now been modified. In summary, the Aβ histidine residues function as binding ligands for all specifically interacting metal ions studied (Paper I-V).

4.1.2 Metal ion binding in membrane-mimetics Copper and zinc ions in relation to the Aβ peptide have been studied to a high extent by several groups (252, 265, 310, 333, 338–343). In Paper IV we introduced metal binding effects when the Aβ40 peptide is in a membrane mimicking environment. Liposomes modeling a biological membrane in terms of lipids, curvature and size are not suitable for NMR experiments; instead SDS micelles were used as a model system. In a lipid environment the Aβ peptide adopts an α-helical secondary structure. Interestingly, the metal binding properties of the Aβ peptide are not significantly affected by the incorporation into the SDS micelles, explained by the free N-terminal part of the peptide also in a micelle complex.

4.2 Metal-Aβ complexes are unable for incorporation into fibril ends (Paper I) A general feature of metal ions, displaying specific binding towards the Aβ peptide in sub-stoichiometric concentrations relative the peptide concentration, is the ability to interfere with the Aβ aggregation behavior. The rate of amyloid formation is reduced, previously demonstrated with Aβ40 and Zn(II) ions where the elongation rate constant is the one most affected parameter without changing the overall aggregation mechanisms (265). In Paper I we studied two different metal ions, Ag(I) and Zn(II) (265) to further investigate the molecular mechanisms. Both ions bind specifically to the N- terminal part of the Aβ peptide via three histidine residues as the main ligands. First, the approach of using Ag(I) ions to perturb the Aβ fibrillization is discussed, and secondly the Ag(I) results are compared to previously

34 published Zn(II) data (265) for a general description and a proposed model of transition metal ion perturbation of Aβ fibrillization kinetics.

4.2.1 Ag(I) ions attenuate Aβ fibrillization by interfering with fibril-end elongation ThT is not suitable for measurements in the presence of Ag(I) ions. Therefore pFTAA was used instead as a fluorescent probe for detection of amyloid material in Paper I. Ag(I) ions attenuate the bulk Aβ fibrillization kinetics in a concentration-dependent manner with similar amounts of amyloid material reached after the reaction, confirmed with AFM imaging. Bulk experiments only show the average behavior of the fibrillization reaction and sigmoidal curve fitting of the kinetic traces does not tell what kind of processes that underlie the observations. Previous studies reported affected Aβ fibrillization kinetics in the presence of Zn(II) ions without any changes of the dominating processes (265). To shed more light on the mechanistic details we used our data sets for a global fit analysis (119, 133, 192, 200, 201, 208) to describe the fibrillization reaction in terms of microscopic rate constants. Further processing of our data provided insights about the elongation rate constant, k+, as the one most affected by Ag(I) ions, even though the other rate constants kn and k2 also contribute to the fibrillization – but probably to a lower extent. This observation was supported by seeding experiments with fibrillary material added at t=0. The presence of seeds provides many sites of fibril end elongation and secondary nucleation to fibril surfaces – whereas the contribution of primary nucleation is minimal. For a low seed concentration elongation and secondary nucleation are dominant, whereas with a high concentration of seeds the slope of the linear dependence of the initial fluorescent intensity signal corresponds to the k+ multiplied with the concentration of aggregates. The relative k+ both from non-seeded and seeded experiments show a similar dependence on the presence of Ag(I) ions, strongly indicating fibril end elongation as the processes most affected by Ag(I) ions (Paper I). But, several questions still remain. How does Ag(I) influence the fibrillization kinetics and what is the mechanisms behind this observation? Does it originate from the monomeric metal binding or not? Does the metal ion affect the peptides during the lag phase of aggregation, or at later stages? Which conformational stage is the starting point of binding to the metal ion or other peptides?

4.2.2 Fibril growth attenuation originates from Aβ peptides bound in metal complexes The kinetic model in the previous section does not provide conformation/structural information. Instead NMR experiments were used to

35 compensate for the lack of structural data. In Paper I the Ag(I) binding to Aβ40 was characterized in detail. The affinity for Ag(I) ions towards the Aβ peptide is in the µM range, explored with a variety of techniques (NMR HSQC, NMR diffusion, fluorescence). In contrast to all other metal ions studied, except Pt(II) ions (Table 1), Ag(I) ions induce chemical shift differences in addition to line broadening of the NMR signal (Paper I). This binding was further characterized by NMR CPMG relaxation dispersion in terms of structure and dynamics. The metal:Aβ binding was assumed to behave as a two state process with a 1:1 binding mode, with a “free” state and a metal-bound state (the bound state likely includes several different bound populations, here encountered as the average one). From CPMG relaxation dispersion experimental information was obtained about the metal-bound state in terms of the exchange rate between the “free” and the bound state and the population for each residue affected by the metal ion binding. Non-specific residue information was also obtained in terms of Δδ and R2. The exchange rate between the two states is on the millisecond timescale with ~10% of the peptides as metal-bound. The chemical shift changes observed in HSQC spectra correlate with the chemical shifts Δδ derived from the relaxation dispersion experiments, which suggests that we studied the same process and bound state with both experiments. This is an extension of previous Zn(II) data (265), where no chemical shift changes from HSQC spectra are visible to be compared to the ΔδZn, relaxdisp. Binding and folding often comes together, and to further characterize the binding PFG diffusion experiments were performed. From these experiments the diffusion coefficient can be extracted, which in the presence of Ag(I) ions increases, indicative of a metal:Aβ complex that diffuses faster than the apo- Aβ. This behavior becomes clearer by using the Stokes-Einstein relationship of the diffusion coefficient and the hydrodynamic radius. The hydrodynamic radius is here based on the assumption of a sphere and is decreased to 16.3 Å in the presence of Ag(I) compared to the apo-state of 17 Å (Paper I). In other words, the Aβ peptide becomes more compact when coordinating a metal ion. Based on our observations from detailed NMR data of the metal-bound state – how come that just a minor fraction of the available population of Aβ peptides give rise to a clear attenuation of the fibrillization kinetics? To be able to answer this question we scrutinized all the data and calculated the theoretical “free” monomeric concentration, m(0), for each Ag(I) ion concentration for each sample in the fibrillization kinetics experiment based on the affinity for the Ag(I):Aβ complex. The higher Ag(I) concentration, the higher portion of peptides in the bound state → lower available Aβ concentration. Hence a theoretical value of m(0) for each Ag(I) concentration was obtained, and this information together with the experimental kinetics curves (Paper I) were globally fitted. The results revealed a reasonable good fit with a dissociation constant in the same range as the values determined with other techniques. In other words, the Ag(I) ions redirect active Aβ monomers

36 towards a detour on the fibrillization pathway, where the Ag(I):Aβ complex is not available for fibril incorporation. Are these observations only applicable to Ag(I) ions alone, or does it apply for metal ions in general?

4.2.3 Model: Metal ion perturbation of the fibrillization process In Paper I we propose a model of sub-stochiometric transition metal ion perturbation on the Aβ fibrillization kinetics, where structural information of the metal:Aβ complex is proven to be linked to the bulk aggregation kinetics behavior explained by microscopically reduced elongation rates (Figure 10).

Figure 10. Proposed model of the effects on the Aβ peptide amyloid aggregation behavior by interactions with sub-stoichiometric metal ion concentrations. Metal ions with a specific histidine-dependent interaction towards the Aβ peptide attenuate the overall fibrillization. This effect originates from a dynamic metal ion interaction with monomeric peptides, making the monomeric pool of peptides unavailable to be incorporated into the fibrils, described by a reduced elongation rate. PDB ID 2M9S.

This model was developed by comparing the obtained Ag(I) ion data with previous data on Zn(II) (265). The binding of Ag(I) and Zn(II) ions towards the Aβ peptide is not identical but similar for induction of a putative fold and interference with the monomeric Aβ peptide. The different capacities are reflected in a lower concentration of Zn(II) needed for a bound population and subsequent attenuated macroscopic and microscopic processes in the fibrillization kinetics. Both ions present a monomeric interaction with an impact on the later stages of the fibrillization pathway – reduced overall fibrillization kinetics via reduced elongation rate. The differences in retardation efficiency may originate from the differences of dissociation constants, as well as other discrepancies (Table 1). To conclude, histidine- coordinated metal ions can retard the overall Aβ fibrillization kinetics by mainly interfering with the process of fibril end elongation originating from an aggregation-inert metal:Aβ complex, while primary and secondary nucleation processes are not as much affected.

37 4.3 Formation of reactive oxygen species by metal:Aβ complexes (Paper IV) Cells and organisms may be exposed to oxidative stress and several protective systems have been evolved during the evolution to deal with ROS, both as inductively/constitutively expressed proteins like SOD, catalase, and glutathione peroxidases, and dietary supplied antioxidants such as vitamin C and E. Oxidative stress is also used as a defense agent against foreign intruders such as bacteria. However, compromised defense systems or overproduction of oxidative stress in sensitive locations is an issue, especially for tissues as sensitive as the brain. Oxidative stress has been implicated as one route of toxicity in Alzheimer’s disease (344). The Aβ peptide, as many biomolecules, is involved in oxidation/reduction reactions. Post-translational modifications have also been reported following oxidative stress, such as histidine modifications, phosphorylation of the serine residues, and crosslinking between aromatic residues such as the reaction of forming a covalent dityrosine crosslink between two adjacent Aβ peptides (345). Cu(II) ions are redox-active, via redox-cycling with Cu(I) and the Haber- Weiss reaction, and in Paper IV we tested the Cu(II)-facilitated hydrogen peroxide formation over time in a biochemical assay in the absence and presence of Aβ (346–348). Over time, samples of 3 µM Cu(II) ions in buffer generated about 30 µM hydrogen peroxide. In the presence of Aβ the formation of hydrogen peroxide was not altered – in other words, the Cu(II)- binding properties of Aβ is not strong/long-lived enough to hinder hydrogen peroxide formation. Neither was the generation of hydrogen peroxide significantly affected by the presence of SDS micelles. The only experimental conditions able to shift the generation of hydrogen peroxide were apo-SOD1 and the chelator EDTA. EDTA is a strong metal ion chelator. Despite the lack of effect of Aβ in these experiments, the Cu(II)-binding Aβ peptides might be of interest in terms of redirecting redox-active ions in close proximity to the plasma membrane due to the membrane-binding properties. Noteworthy, the murine Aβ peptide binds Cu(II) with higher affinity compared to the human one, with less hydrogen peroxide formed for the murine peptide (326).

4.4 Outlook Metal ions that display a specific binding to Aβ can be used as modulators of the fibrillization process. In Paper I this strategy was addressed, and it remains to be verified if it can be translated onto other amyloidogenic and misfolding proteins as well, since several misfolding proteins implicated in misfolding protein diseases display metal binding properties. Further careful analysis of other relevant metal ions to be put into the model proposed in

38 Paper I is ongoing. Beyond metal ions, additional efforts to explain the microscopic and macroscopic effects on both aggregation and fibrillization kinetics are highly endorsed for other modulators too. The metal binding to Aβ is different compared to metalloproteins where the metal binding site is highly specific. Is there a secondary binding site occupied at high metal ion concentrations accompanied with disruption of key interactions needed for amyloid formation? Both intra- and intermolecular exchange between the metal ion and the Aβ peptide is possible. In Paper I the metal-bound populated state is determined to ~10% of the total Aβ peptide population with an exchange rate on the millisecond time scale. The low population, the relatively low Aβ concentration used (in terms of NMR spectroscopy), together with a transient metal ion binding contribute to a challenge to obtain a structure(s) of the metal bound state(s). Optimal experimental conditions for stabilization of the bound state would be very much appreciated in the field, such as forcing the bound population into a more stable state. The relation of the metal ion induced perturbation of fibrillization and cytotoxicity is still an unsolved question. The findings from my in vitro thesis work cannot tell anything about potential toxicity (170). Notably, misfolded proteins with exposed hydrophobic patches can be toxic, independently of metal ions, and sole metal ions are cytotoxic in a dose-dependent manner. Whether toxicity by metal-induced protein structures is additive, synergistic or antagonistic is dependent on the metal ion, and details still have to be elucidated. Homeostasis of endogenous metal ions is tightly regulated but may be disturbed by contaminants. Competition between endogenous metal ions with metal ion contaminants may be biologically relevant. Formation of ROS during the aggregation and potential toxic endpoints are further points of interest. One of the main questions in the field is about multimeric structures of Aβ peptides, and how such structures proceed to larger and insoluble fibril structures – or the other way around with fibril and fibril-like structures anticipating oligomers as a source, or acting as catalytic sites for oligomeric peptide formation. Related to metal ion chemistry, the roles of metal ions in the growth or formation of such structures, regulation of metal balance and AβPP processing, and the functionality of Aβ-degrading proteins, are important questions to address. The relevance of metal binding towards the Aβ peptide in health and disease remains to be elucidated. The relatively low Aβ affinity for metal ions does not speak for a significant metal ion binding role during physiological concentrations, in the presence of i.e. copper proteins with substantial higher affinity for copper ions. On the other hand, both metal ions (Cu and Zn) and Aβ concentrations are increased during synaptic activity (100, 277, 278). Iron overload is as well an important issue related to advanced age and Alzheimer’s disease (79, 288).

39 5. Small molecules and peptide construct interactions with Aβ peptide assemblies

“Being a scientist is a special privilege: for it brings the opportunity to be creative, the passionate quest for answers to nature's most precious secrets, and the warm friendships of many valued colleagues.” Stanley B. Prusiner

5.1 Hydrophobic compounds found in cigarette smoke (Paper V)

Alzheimer’s disease is more abundant in a cigarette smoking sample population compared to a non-smoking sample population (349). To study the effects of cigarette compounds on the Aβ peptide from a molecular perspective, we studied a selection of polycyclic hydrocarbons (PAHs) and nicotine present in cigarettes and cigarette smoke in Paper V. All molecules except nicotine promoted the Aβ aggregation kinetics under our experimental conditions. Interestingly, no specific binding studied with 2D NMR HSQC was found for these compounds, yet the aggregation kinetics were affected, possibly by interactions with more ordered structures containing Aβ peptide multimers. Such sizes of Aβ assemblies are invisible by solution NMR and may explain the lack of observed specific interactions. To conclude, small hydrophobic compounds found in cigarette smoke promote the Aβ aggregation without a specific interaction with the monomeric Aβ peptide. Hydrophobic and non-specific interactions and when/where on the fibrillization pathway a detour from the original path is initiated need to be further studied, and not only limited to molecules in cigarettes but also other hydrophobic compounds are of interest.

5.2 Designed peptide constructs (Paper VI) One therapeutic strategy used in clinical trials to combat Alzheimer’s disease is to target misfolded Aβ peptides distinguishable from both Aβ monomers and Aβ fibrils. The blood brain barrier is undoubtedly a hinder of passive immunization therapeutics administered orally or intravenously, due to the limitations of low bioavailability. One efficient strategy is to use drug cargos attached to cell-penetrating peptides (CPPs) able to cross the blood brain barrier for effective delivery into the brain. CPPs occur both endogenously and as engineered peptides. The cell-penetrating abilities of certain proteins were discovered in the 1980s and since then the field has

40 evolved with an increased understanding and steps into clinical applications (350). CPPs are non-toxic peptides in the size range of 5-30 amino acids long and give rise to protein delivery via cellular uptake. The process can be both energy-dependent and energy-independent (351). The cell-penetrating properties open up an avenue of potential applications to deliver drugs of various kind (352). This approach has also been utilized in amyloid research (353). Previous research studied prion protein (PrP) infected cells and inhibition of the progression of prion infection by different proteins with cell-penetrating properties (354, 355). The strategy was based on steric hindrance of the pathologic conversion of normal PrPC to the so-called scrapie form PrPSc evolved by the N-terminal part of the murine prion protein, mPrP1-28. This part of the protein consists of a signal sequence (mPrP1-22) and a positively charged hexapeptide (mPrP23-28). The mPrP23-28 motif has been reported to have a high affinity towards the PrPSc version of PrP (356), possibly explaining the observed behavior. Due to the cell-penetrating properties the mPrP1-28 peptide was internalized inside the cells and able to effectively attenuate the prion infection (357). Interestingly, another signal sequence for secretion from the first 19 residues of another protein, the neural cell adhesion molecule – 1 (NCAM) conjugated with the hexapeptide from the PrP (mPrP23-28) was even more efficient to inhibit prion infection. The signal sequence targeting for secretion seems to be important for the observed effect, with properties of hydrophobic segments compared to CPPs in general. The physiological function of the native NCAM protein is related to adhesion mechanisms between neurons, neuronal growth and neurite fasciculation, and located in the plasma membrane. The NCAM protein does also potentially interact with PrP and AβPP (358–360). Does this inhibitory effect on misfolding prion progression also apply for other amyloidogenic protein processes? This question was addressed in vitro and in cells in Paper VI. Two different peptide constructs were studied in Paper VI, one identical sequence from the previous study (357) called NCAM-PrP (MLRTKDLIWT LFFLGTAVSKKRPKP-NH2), whereas a second one with the cationic hexapeptide based on a part of the Aβ sequence (NCAM-Aβ) was also added as a proof of principle peptide construct. In the scope of this thesis only the NCAM-PrP variant is discussed. The NCAM-PrP peptide construct is predominately unstructured in aqueous solution and forms α-helical secondary structures in 30% HFIP (Paper VI) and in membrane mimetics such as SDS micelles. The peptide construct is stable in pure water at low pH for several weeks but incubated in buffer solution at physiological pH the NCAM-PrP ([μM]) starts to aggregate/precipitate. In Paper VI the NCAM-PrP peptide construct efficiently disrupts the typical sigmoidal ThT fluorescence curves of Aβ42 peptides in a ThT aggregation kinetics assay. At a ratio of 0.5:1 (NCAM-PrP:Aβ) the ThT kinetic trace shows an instant but small increase in ThT fluorescence that

41 levels out to a stable, low signal fluorescence intensity level. The low ThT fluorescence intensity is indicative of a low level of ThT-active aggregates/fibrils, and transmission electron microscopy (TEM) images confirmed the absence of typical amyloid fibrils. These observations indicate that the Aβ42 fibrillization is inhibited in the presence of the NCAM-PrP peptide construct. The mechanism behind this effect was further investigated with solution 1D and 2D HSQC NMR experiments. The impact of NCAM- PrP on monomeric 15N-Aβ peptides was investigated and reported as non- specific interactions; however, a gradual and uniform loss of signal was observed in the spectra together with a clear precipitation visible by the naked eye. To avoid precipitation and to constrain the 15N-Aβ peptides in a monomeric state, the same experiment was repeated in SDS micelles (unpublished data). Both monomeric 15N-Aβ peptides and NCAM-PrP construct peptides adopt α-helical secondary structures located in SDS micelles, however no immediate interaction was observed. We further examined the impact of the NCAM-PrP peptide construct on 15N-Aβ peptides over time (unpublished results). 15N-Aβ peptides were incubated in the presence and absence of sub-stoichiometric concentrations of NCAM-PrP for 48 hours and a 1H-15N-HSQC spectrum was recorded before and after the incubation. As expected, the sample of 15N-Aβ peptides alone started to aggregate during the incubation proved by reduced signal resonance intensity. Surprisingly, the 15N-Aβ sample supplemented and incubated with NCAM- PrP showed a less pronounced signal loss compared to the sample without (Figure 11). Taken together, these observations suggest a weak, if any, interaction with monomeric 15N-Aβ species, and with non-identified interactions with more ordered structures invisible by NMR. Another aspect is potential dissociation of ordered structures to monomeric species as part of the explanation of maintaining the 15N-Aβ sample monomeric to a higher extent in the presence of NCAM-PrP, this is also supported by the observations from the ThT assay and TEM imaging (Paper VI). The inhibitory effect of NCAM-PrP on the Aβ42 fibrillization kinetics is hence not solely explained by an interaction with monomeric Aβ species. The Aβ-NCAM-PrP interaction and effects on the Aβ peptide aggregation behavior were further investigated with ELISA and dot blot assays using mono- and polyclonal antibodies recognizing oligomeric Aβ structures (Paper VI). The signal of oligomers increased linearly when Aβ was incubated over time in the absence of the NCAM-PrP, followed by a rapid decrease currently proposed as a shift of the oligomeric population towards fibrils. In contrast, Aβ incubated in the presence of equimolar concentrations of NCAM-PrP decreased the signal of oligomeric structures by 35-50%. The NCAM-PrP peptide construct thus interferes with the formation of oligomeric Aβ species in this assay. The inhibitory effects of Aβ fibrillization in vitro in a simple model system is one isolated aspect, and effects in cellular systems is another one. The

Figure 11. Aβ and NCAM-PrP peptide construct interactions. (A) 2D NMR 1H-15N-HSQC 15 spectra of 20 μM N-Aβ40 before (dark blue) and after 48 hours incubation (light blue). Below are the relative intensities plotted against the Aβ primary amino acid sequence shown. In (B) the measurements in (A) were repeated but in the presence of 5 μM NCAM-PrP (dark blue) and after 48 hours incubation (green). (C) ThT fibrillization kinetics experiments. 5 μM Aβ42 peptides were incubated in 10 mM MOPS buffer at physiological pH +37 °C supplemented with 10 μM ThT, in the presence and absence of 10% seeds and 5 μM NCAM-PrP peptide constructs.

NCAM-PrP construct peptide was also tested in N2a mouse neuroblastoma cells (Paper VI). Interestingly, the NCAM-PrP peptide was able to rescue the N2a cells completely from the Aβ42-induced toxicity in a cell viability assay

43 where the cells were supplemented with the two peptides simultaneously. Cells were also exposed to fluorescently labeled peptides and studied in confocal fluorescence microscopy imaging after 24 hours of incubation. The peptides were co-localized in mitochondria and lysosomes, observations that indicated internalization of both peptides and localization to similar sites in the cell. In addition, cells were also pre-treated with Aβ peptides for 24 hours followed by supplementation of NCAM-PrP peptide constructs. It turned out that the NCAM-PrP peptides were able to target intracellular Aβ peptides, visualized by confocal fluorescence microscopy imaging and cell viability measurements (Paper VI). The NCAM-PrP concentration used together with Aβ was also tested to distinguish any toxicity of the construct alone and the cell viability for those samples were close to the background level. In summary, the NCAM-PrP peptide construct rescued cells from Aβ-induced neurotoxicity. The designed peptide construct inhibits Aβ42 fibrillization (ThT assay and TEM imaging), prevents formation of oligomeric species (ELISA and dot blot assay) and rescues Aβ42-induced neurotoxicity and targets both intracellular and extracellular Aβ42 (Paper VI). NMR experiments proposed that the involved interactions between Aβ and NCAM-PrP does not originate from monomeric Aβ peptides. So, if it is not monomeric interactions, what kind of structural interaction lies behind the inhibitory effects of NCAM-PrP? To enlighten this question seeding experiments were added to the cassette of ThT kinetics experiments, for an attempt to identify secondary nucleation pathways from primary nucleation processes. Pre-formed seeds at a high seed concentration was added at time zero to monomeric Aβ42 peptides in the presence of different concentrations of NCAM-PrP peptides (unpublished results). In the absence of NCAM-PrP, the Aβ42 peptides seeded with pre- formed aggregates reached the plateau phase before the samples in the absence of seeds left the lag phase, indicative of dominating secondary nucleation processes. Intriguingly, in the presence of NCAM-PrP the effect of pre-formed seeds was abolished (Figure 11). In other words, despite an increased number of fibril surfaces and fibril ends, the growth of aggregate mass is similar to the conditions without seeds. These findings need to be further verified, but the secondary nucleation processes may be hampered by NCAM-PrP, along with potential additional processes. In summary, the NCAM-PrP peptide construct with a hydrophobic signal sequence + a polycationic sequence inhibits and detouring the Aβ aggregation in vitro with reduced amounts of oligomers and fibrils, and attenuated neurotoxicity effects in cellular model systems. The NCAM-PrP peptide construct stabilized the Aβ peptide in a non-amyloid state, that may be a promising approach for future therapeutic strategies. Obviously, further development of the CPP-based peptide strategy is needed, and more information about the interaction, optimization, D-variants, and other sequence improvements are yet to come in ongoing projects.

44 6. Two interacting proteins implicated in Alzheimer’s disease – Aβ and Tau

“Were always together, were one of a kind, three words describes us “partners in crime”.” Unknown

There are several hypotheses regarding Alzheimer’s disease, and two of them are the amyloid cascade hypothesis and the Tau hypothesis (2, 76, 77, 361). These two hypotheses occur both individually and combined. Common denominators of Aβ and Tau are their properties of forming amyloid and paired helical filament structures from disordered monomeric structures. The Aβ peptide and Tau protein (362, 363) are two proteins with highly disperse properties and functions (Figure 12). Tau protein is about ten times larger than Aβ, with a net charge of +3 compared to -3 for the Aβ peptide under physiological conditions. The physiological roles of the two proteins are different. Detailed Aβ peptide information is presented in section 2.4. Tau protein has a distinct native function of promoting axonal microtubule assembly and microtubule stability and belongs to the microtubule-associated protein (MAP) family (364). There are six isoforms of Tau, and the largest one consists of 441 residues (365). Tau protein is highly unstructured with a low content of secondary structures, but with regions exhibiting β-sheet propensity (365). The affinity towards microtubules decreases upon phosphorylation (362, 366) and is part of the equilibrium between association and dissociation of microtubules. In Alzheimer’s disease, Tau proteins are found hyperphosphorylated and aggregated into helical filaments (76). In vitro, native Tau protein does not aggregate into ThT-active aggregates unless incubated in the presence of negatively charged polyions such as heparin (367). Interplay between Aβ and Tau protein including co-existence and combined pathological effects has been reported (76, 146, 368–382). It has been proposed that Aβ amyloidosis induces or enhances Tau pathology (76, 146, 368–381), and Aβ toxicity has been described as Tau-dependent (368). The combined pathological effects are partly seen as synergistic effects from both Aβ and Tau (76, 146, 368–381). Aβ/AβPP may trigger Tau aggregation, and Tau protein may trigger Aβ aggregation. Tau deficiency, Tau-/- models, are not susceptible to Aβ toxicity (76, 146, 368–381). Transgenic mice overexpressing both AβPP and presenilin-1 were reported with lower burden of senile plaques when Tau was suppressed (368). If Aβ is mainly an extracellular peptide and Tau is an intracellular microtubule-binding protein – can these proteins make contact in vivo? However, both Aβ- and Tau

Figure 12. Native Tau protein. (A) Tau protein belongs to the microtubule-associated protein (MAP) family and stabilizes microtubules in the cell, Tau is here visualized in dark blue colour from PDB ID 6CVN (383). (B) The longest isoform of Tau protein is 441 residues long and the microtubule-binding domains are designated as R1, R2, R3, and R4 (363).

aggregates exhibit prion-like transmission properties for spreading from one neuron to another, observed both in vitro and in patients (384–386). Therefore the Aβ peptide is not only limited to the extracellular space, but is rather also present in intracellular compartments (165). Under physiological conditions the Aβ concentration in the human brain is low (nanomolar), but may be higher in the μM range in specific cellular compartments (96, 108). The physiological concentration of Tau protein is about 2 μM (387, 388). The in vivo estimated concentrations and the ability for both proteins to propagate from one cell to another actually provide reasonable opportunities for the two proteins implicated in Alzheimer’s disease to physically meet. Paper VII investigated the interplay between the Aβ40 peptide and native full-length Tau protein on the molecular level in vitro and how Tau influences the aggregation properties of the Aβ peptide.

46 6.1 Full-length native Tau protein prevents Aβ fibrillization (Paper VII)

In Paper VII we studied the impact of Tau protein on the Aβ40 peptide fibrillization. From our findings, a simple schematic model of interaction stages during the Aβ fibrillization pathway is suggested and highlighted in Figure 13. We monitored the fibrillization of Aβ40 with a ThT assay in the presence of increasing concentrations of Tau proteins. Whereas the Aβ40 peptide sample shows a typical ThT fluorescence signal curve with a sigmoidal shape, Tau protein prevents the Aβ fibrillization process in a concentration-dependent manner observed as increased aggregation halftimes and decreased fluorescence intensity at the end of the experimental time. Based on these observations, Tau protein is proposed to prevent the Aβ fibrillization processes. However, this analysis does not tell about the underlying microscopic nucleation mechanisms, nor at what stage and how the fibrillization is attenuated. The ThT molecule used in the ThT assay does not recognize amorphous aggregates and to investigate the structural state at the end of the Aβ aggregation in the presence of Tau protein, TEM imaging was used. Compared to the 10 µM Aβ alone sample that formed characteristic fibrils, in samples with Aβ in the presence of 2.5 or 10 µM Tau proteins only non- fibrillary structures were present. These findings hence validate the observation from the ThT assay with less ThT-active structures in the presence of Tau. For more detailed information, monomeric Aβ was studied with high- resolution NMR spectroscopy. 2D NMR 1H-15N-HSQC spectra of 10 µM 15N- Aβ in the absence and presence of equimolar concentration of Tau were recorded before and after 15 hours of incubation. All samples showed monomeric 15N-Aβ resonance signals with similar signal intensities for both samples before the incubation started, and after the incubation the 15N-Aβ sample without Tau protein lost all signals down to the noise level due to aggregation into larger structures. In contrast, the sample co-incubated with Tau protein still showed monomeric 15N-Aβ resonance signals, indicative of remaining monomers in solution. Based on these observations, it is relevant to propose that the key Tau interaction explaining the attenuation of the Aβ fibrillization is not with monomeric Aβ – rather with more ordered structures such as oligomers and fibrils (Figure 13). To test this hypothesis, we performed fluorescence polarization experiments to measure the affinity between Tau and Aβ monomeric-, oligomeric-, and fibril structures. So called oligomeric structures were generated by incubation of monomeric Aβ peptides at certain time points. Interestingly, the fluorescence polarization experiments confirmed a low affinity towards monomeric Aβ (KD >100 µM), but with higher affinity for oligomeric species (KD of ~10 µM) and the highest affinity for fibrils (KD of ~1 µM). Taken together, the higher affinity for larger structures (oligomers,

Figure 13. Full-length, native Tau protein prevents Aβ fibrillization. Tau influences the Aβ fibrillization at several stages, with a range of fibrils > intermediate structures > monomeric species, determined by various biophysical techniques.

fibrils) compared to monomers and an increased monomeric population in the presence of Tau protein is indicative of Tau interference at several levels related to the Aβ fibrillization: both by binding to the fibril surface preventing formation of new aggregates on the surface or elongation to the fibril ends, and by facilitating dissociation of Aβ oligomeric structures into monomers.

6.2 Outlook Two interacting proteins very much entangled in Alzheimer’s disease are intriguing and of interest for further investigation. Are the Aβ peptides and Tau proteins partners in crime, or a result of dysregulated biochemical processes? The clinical implications of such interactions remain to be tested/verified, and multifaceted therapeutic interventions towards both Aβ species and Tau protein are suggested. Detailed structural information of the Tau-Aβ interaction coupled to biological/physiological conditions is highly desirable. We first wish to investigate which domain of Tau protein is responsible for the prevention and detouring of the Aβ fibrillization. A working hypothesis interesting to investigate is if the absence of fibrillated Aβ species inside cells might be related to the presence of native full-length Tau protein.

48 7. Concluding remarks and future perspectives

“Always make a total effort, even when the odds are against you.” Arnold Palmer – ”Athlete of the Decade” in the 1960s

“Spela bollen som den ligger och banan som den är” Regel 13 i golfboken

A detour of amyloid building blocks can be achieved by perturbing the fibrillization processes by interacting metal ions and molecules. The Aβ peptide interacts with various ions and molecules, both at the monomeric stage and as larger assemblies. In brief, based on our findings we specifically conclude that:

- From a total of 33 metal ions (27 ions studied within this thesis), 12 metal ions interact specifically with the Aβ peptide. - Based on the observations and the Pearson acid-base concept (336), the Aβ peptide interacts with intermediate-soft electron acceptors, both monovalent and divalent metal ions. Notably, not all divalent metal ions interact specifically with the Aβ peptide.

- Specific metal ion interactions to the Aβ40 peptide with an affinity high enough modulate the Aβ peptide self-assembly. At high metal ion concentrations ionic strength effects are present. - A common interaction mechanism appears for monovalent Ag(I) and divalent Zn(II) ions with Aβ peptides. - The pool of available Aβ monomeric peptides is reduced once Ag(I) and Zn(II) ions are bound to the Aβ peptide. This interaction results in an overall attenuating effect of Aβ fibril formation linked to a reduction of the elongation rate. - SDS-bound Aβ peptide binds Cu(II) ions and copper ions generate hydrogen peroxide, implicated as a source of ROS and oxidative stress. - Small hydrophobic compounds, PAHs, present in cigarette smoke promote the Aβ amyloid aggregation kinetics. - Designed CPP peptide constructs, NCAM-PrP peptides, comprised of a hydrophobic signal sequence targeting for secretion and a polycationic

hexapeptide affect oligomeric structures of Aβ multimers, and inhibits Aβ42 fibrillization.

- In cells, NCAM-PrP peptides rescue Aβ42-induced neurotoxicity, and targets

both intracellular and extracellular Aβ42. - Soluble, full-length, and non-phosphorylated Tau-441 protein prevents Aβ aggregation. The Aβ fibrillization process is not prevented by Tau interaction with Aβ monomeric species, but rather with fibrils and oligomeric species of Aβ.

49 7.1 Outlook There are still many remaining questions. The Aβ amyloid aggregation processes are heterogenous, and easily affected by external factors. The interplay between metal ions and the Aβ peptide in relation to cytotoxicity, mitochondrial dependence, competitive binding and more importantly, the biological and pathological relevance, are a few areas of further possibly investigation. Are these NMR and aggregation kinetics studies applicable to other proteins as well? The interactions and aggregation modulators can be used both to learn more about the underlying fibrillization processes and for development of potential and promising therapeutic strategies. CPP-based inhibitors of Aβ fibrillization might be a promising therapeutic possibility. A body of different ions and molecules (endogenous, exogenous, and disease-relevant ones) interacts with the Aβ peptide and changes the aggregation processes on the microscopic level. By studies of such interactions, clues about the aggregation mechanisms can be obtained as fundaments for further studies to design better, sustainable, and suitable aggregation modulating substances to therapeutically target amyloid-forming proteins in diseases. Is it an advantage with slow fibrillization or is fast fibrillization more beneficial if translated into cellular systems? It is not a simple question with a simple yes/no answer based on the behavior of aggregation in bulk experiments. Slow aggregation may contribute to a longer time window with Aβ peptides in an oligomeric state before the fibril state is reached, whereas fast aggregation kinetics may contribute to aggregation of species not readily available for aggregation in the first place. A decrease of aggregation- competent units such as the monomer concentration, nuclei, certain oligomeric structures, or available fibril surfaces will certainly affect the aggregation kinetics for sure. On the other hand, the nucleation processes may be affected directing the aggregation pathway to a different path. To combat protein misfolding linked to the pathogenesis of neurodegenerative diseases further steps still need to be taken, likely facilitated with substantial resources and collaborations. Preventative actions are important, long-term goals and ambitions may be crucial for successful treatments in misfolding protein diseases. Detour and an escape from the fibrillization pathway may be one step further.

50 8. Populärvetenskaplig sammanfattning

”Allt bör göras så enkelt som möjligt, men inte enklare” är ett citat myntat av Albert Einstein. Projektet i den här avhandlingen använder ett enkelt modellsystem för att studera komplexa processer underliggande Alzheimers sjukdom enligt den så kallade amyloid-hypotesen.

Bakgrund Människan består utav 37,2 biljoner (1012) celler, där varje cell är i genomsnitt cirka 100 mikrometer = 0,0001 meter = 0,1 millimeter i diameter. Om varje cell läggs på en enda lång rad så räcker detta band av celler till hela 93 varv runt jordklotet, motsvarande fem tur och returresor till månen. Enskilda celler (n=1) kan ingående studeras med modern teknologi och vetenskapen idag har gjort det allt mer vedertaget att studera enskilda molekyler i laboratoriet. Ritningen för en människa finns tätt packad i en lösenordskyddad kärna i varje cell som ett genom bestående av DNA. DNA kodar för olika ”arbetsmyror”, även kallade proteiner, som behövs för att sköta processer och händelser i en cell. Det finns på ett ungefär 20-25 000 gener som i sin tur kodar för cirka 100 000 olika proteiner. Ett protein, en molekylär maskin, är en polymer av sammanlänkande aminosyror, vars ordning och typ avgör proteinets struktur och dess funktion. Myosin och aktin är två exempel på typiska proteiner i muskelvävnad som med generella termer ofta associeras med protein i det vardagliga livet. Människokroppen består av många olika sorters proteiner med livsviktiga funktioner såsom sockerupptag (insulin), signalering (hormoner, signalsubstanser), energi-produktion i cellens kraftverk (proteiner i mitokondrier) och cellskelett (aktin, keratin, tubulin). Hjärnan består av en mosaik av flera olika sorters celler. En av dessa celltyper, neuroner, består av en cellkropp (soma), en lång axon för transport av nervimpulser samt dendriter som samlar upp och delar vidare kemisk information. Alzheimers sjukdom, en av de vanligaste demenssjukdomarna, kännetecknas av att hjärnan förtvinar genom förstörelse av neuroner (se Figur 14). Förtvingen leder till en påverkan på vitala funktioner samt hjärnans kommunikation mellan nervcellerna, för att slutgiltigt helt upphöra. Denna neurodegenerering tros vara en följd eller orsak av aggregerade (”ihop- kloggade”) proteiner. Aggregerade proteiner förlorar både sin naturliga funktion och är svåra att bryta ned för cellen, samt klibbar ihop andra

51 livsnödvändiga system som i sin tur bidrar till ett tillstånd som inte är optimalt för cellens överlevnad och livskvalité. Flera sjukdomar är associerade till neurodegeneration, såsom Alzheimers sjukdom, Parkinsons sjukdom, ALS, Huntingtons sjukdom, Creutzfeldt-Jakobs sjukdom (”galna kosjukan”) samt ”Skellefteå-sjukan”. Proteiner som ofta förknippas med Alzheimers sjukdom är Aβ-peptiden och Tau (se Figur 14). Aβ-peptiden består av både klibbiga och vattenlösliga segment och dess naturliga funktion är ännu inte helt klarlagd. Enstaka och ett fåtal Aβ-peptider är vattenlösliga, men vid start av dess aggregering via olika mellantillstånd sjunker lösligheten tills dess att sluttillståndet av stabila och olösliga aggregat har uppnåtts. Tau har däremot en tydlig funktion i att stabilisera de protein som i sin tur stabiliserar cellens ryggrad/skelett. Både Aβ och Tau tenderar att aggregera ihop till proteinaggregat vid sjukdom, både var och en för sig samt i en ömsesidig process.

Figur 14. Typiska kännetecken för Alzheimers sjukdom inkluderar förtvinad hjärnvävnad, celldöd och aggregat av Aβ-peptider och Tau protein. Tau hjälper till att stabilisera cellens ryggrad och axon. I den här avhandlingen har Aβ-peptidens aggregationsförlopp (=processen från enstaka och lösliga peptider till stabila fibriller/aggregat bestående av många peptider) studerats i detalj.

52 Problemställning För att beskriva och förhindra neurodegenerativa sjukdomar är det av största vikt att förstå i detalj hur processen som bidrar till att proteiner aggregerar fungerar, vilket är en del av den här avhandlingen. Här har ett (n=1) protein studerats, den så kallade Aβ-peptiden. Aβ-peptiden har en stark tendens att aggregera – det vill säga att under rätt betingelser så sker aggregation och bildande av kemiskt stabila aggregat spontant. Aggregationsförloppet i samband med olika mellantillstånd anses vara sjukdomsförkallande (=amyloid-hypotesen). Arbetet har inriktats på att studera olika interaktioner i relation till dess effekt på aggregationsförloppet.

Metodik I provrör har molekylära interaktioner och aggregationsprocesser studerats med hjälp av biofysiska metoder — spektroskopi och mikroskopi. För vissa mätningar har även levande celler, neuroner, använts. Aβ-peptiden är en av huvudaktörerna inom amyloid-hypotesen. Detaljstudier har utförts av Aβ- peptidens interaktioner med andra molekyler (metalljoner, designade peptider och Tau protein) och dess effekt på aggregationsförloppet.

Resultat och slutsatser Aβ-peptiden påverkas i mångt och mycket av flera olika faktorer och beroende på vilken typ av interaktion som föreligger kan effekten te sig olika ut. Resultaten tyder på att enskilda Aβ-peptider (n=1) gärna binder svagt till flera olika metalljoner. Den här typen av metallbindning påverkar Aβ- peptiden så pass mycket att det totala aggregationsförloppet tar längre tid till att fullborda aggregationen till olösliga och strukturerade aggregat. Dock får inte alla sorters metalljoner ut den här effekten. Det som krävs är en specifik bindning som är tillräckligt kraftfull och ger en veckningsform som kan leda bort enskilda Aβ-peptider från det spontana aggregationsförloppet. Denna process att Aβ-peptiden tar en omväg innan de hittar rätt i aggregationen påverkar främst tillväxten av befintliga aggregat. Till skillnad från metallbindning till enskilda Aβ-peptider så har andra molekyler en annan typ av effekt för aggregat med många Aβ-peptider (n=~2- 100). Designade peptidsekvenser som har egenskaper för att kunna passera från en cell till en annan kan även påverka Aβ-peptidens aggregationsförlopp. Effekten kommer inte från en interaktion med enskilda Aβ-peptider, utan snarare med multistrukturer av Aβ-peptider som kan variera i form och storlek och även i ytegenskaper. Dessa designade peptidsekvenser samverkar med dessa multistrukturer och förhindrar fortsatt aggregation till fullväxta aggregat samt hindrar även toxiciteten hos dessa multistrukturaggregat. Ett annat protein, Tau, påverkar också Aβ-peptidens aggregationsförlopp genom att samverka med multistrukturer snare än med enskilda Aβ-peptider. Tau förhindrar att Aβ-peptiden bildar strukturerade aggregat, samt bidrar till

53 en ökad fraktion enskilda Aβ-peptider från en aggregatlösning. Det är inte helt klarlagt än, vad den upplösande effekten beror på, men den är tydlig. Tre olika typer av modulatorer (=metalljoner, designade peptider och Tau) av Aβ-peptidens aggregationsförlopp har studerats och de påverkar Aβ- peptiden på olika aggregationsnivåer. Det modulatorerna har gemensamt är att den framkallade störningen av aggregationsförloppet både kan användas för att lära sig mer om de bakomliggande molekylära mekanismerna för aggregation samt ge idéer på möjliga terapeutiska strategier.

54 9. Acknowledgements

”In science, one should use all available resources to solve difficult problems. One of our most powerful resources is the insight of our colleagues”. Peter Agre

If I would list everyone who has been important for and during this thesis work, this section of the thesis would have been longer than the kappa itself. Thereby I will keep it short and simple (without mutual order).

I wish to express my sincere gratitude to my supervisor Astrid Gräslund. Thank you for infinite support, scientific guidance and patience. Thank you for four and a half years of research joy and happy memories. Thank you for being encouraging and a good role model, for accepting me as a PhD student, for giving me freedom to suggest and pursue new ideas and research projects, always having your door open for questions and discussions, and for letting me take part of our global community with collaborators all over the world. It has been an interesting journey. I am grateful for financial support from foundations/organizations making it possible to participate and to present our research on international scientific conferences, summer school courses, and research visits. Thanks to the Biophysical Society travel award, Klas-Bertil and Margareta Augustinssons stipendiefond, the Jubilee donation, K & A Wallenberg Foundation, the Nobel Foundation, the Paul Scherrer Institute, and the Stockholm University Donation Scholarship. My co-supervisor Andreas Barth, thank you for being incredibly supportive for each stage of the PhD process. Thank you for all your help, for being encouraging, for your guidance, and for inspiring me to learn more about IR spectroscopy. For this I am enormously grateful. Extra supervisors, Per Roos and Sebastian Wärmländer, thank you for support, guidance, and for the opportunity to study these incredible interesting research topics. Per, I highly appreciate your enthusiasm and for introducing the field of neurodegeneration from the medical point of view to me, including meetings with patients. Sebastian, thank you for providing many different aspects of protein chemistry/methodology, and for the help to present our research at conferences and to visit other laboratories of our collaborators. I would like to acknowledge Axel Abelein, thank you for sharing your knowledge and your enthusiasm about research so generously. It has been a privilege to be part of our collaborative and joint research projects, and I have learned so much during the process. I highly enjoyed working with our projects. Thank you for excellent feedback and for help with the thesis work.

55 I would like to express my gratitude to Jens Danielsson for your consistent, skilful support and highly valuable input throughout the research projects. Thank you for providing multiple perspectives and for encouraging me to never leave no stone unturned to be able to interpret all the data obtained. ”Stick to the data” will for always be remembered. I wish to show my deepest gratitude to Jinghui Luo, who I have had the fortune to work together with on multiple projects. Thank you very much for all support, scientific advices, and excellent supervision. I have learned very much, all from basic knowledge to new and innovative methods and strategies to study amyloid-related questions and beyond. I am incredibly grateful for the opportunity to visit your lab and group in Switzerland for exciting research projects and modern research facilities, we did accomplish a lot of work in only two weeks! Jüri Jarvet, thank you for always sharing your knowledge and for providing straight answers to all my questions. There have been many interesting discussions over the years. I am forever grateful for everything I learned from you. Thank you for teaching me about FCS measurements. Britt-Marie Olsson, thank you for being the kindest person, always being so happy and friendly. I wish to address the importance of your assistance that has been invaluable in this process. Thank you! I would like to pay my special regards to Göran Eriksson, thank you for all of your help and for sharing your knowledge and beautiful paintings. Thank you for being both a good colleague and for your invaluable friendship. Henrik Biverstål, thank you very much for all your help in the lab and for insightful and enlightening discussions. Sabrina Sholts, thank you very much for being so friendly, welcoming and for accepting me as an internship student for a few weeks in your lab in Washington DC. By far one of the best experiences, to work and to learn new things in an American lab setting with questions concerning bones, metals, ancient water bottles and health. Furthermore, my warmest and deepest gratitude to everyone, who showed an interest in our research, did ask questions on conferences/meetings, questioned our work, supported us with constructive criticism, wrote constructive review comments to improve our papers, and provided reality checks for new perspectives. I am incredibly grateful for the opportunity of being part of the chemical- and amyloid communities for a few years. I would like to acknowledge all past and present lab members of the Gräslund lab that I had the opportunity to meet and work with. Thank you very much for all these years. Thanks to all members of DBB, including the administrative and technical personnel who are invaluable for us working in the lab. Thank you. I wish to pay my special regards to Ann-Britt Rönnell, Alex Tuuling, Erik Sjölund and Matthew Bennett, and former employees Haidi and Torbjörn Astlind, for your endless support and persistent help. I

56 would like to express my gratitude and appreciation to the academic crew at DBB. It truly has been a wonderful/interesting time. Ida Nyqvist and Johan Berg, what would I have done without you when we were facing our first year as lab assistants? You made the teaching responsibility highly enjoyable and you also inspired me with your technical teaching skills and problem-solving approaches. As a plus, or really the most important thing, the students were happy too. I also want to express my gratitude towards Agneta Norén, the great course leader of the course. Thank you - Teamwork makes dreamwork. A sincere appreciation and deepest gratitude to our collaborators over the years, with special thanks to Mazin Magzoub, Elzbieta Glaser, Pedro Teixeira, Salma Al-Adwani, Peter Bergman, Anders Olofsson, Oleg Antzutkin, Lynn Kamerlin, Birgit Strodel, Jan Pieter Abrahams, Yashraj Kulkarni, Qinghua Liao, Ludmilla Morozova-Roche, Igor Iashchishyn, Jonathan Pansieri, Istvan Horvath, Guy Lippens, Shai Rahimipour, Leopold Ilag, and Alex Perálvarez- Marín. Current and former colleagues, friends, and inspirers/motivators: Alexandra Toth, Andreas Carlström, Ann Tiiman, Axel Leppert, Biao Fu, Britt-Marie Sjöberg, Candan Ariöz, Carmela Vazquez Calvo, Cecilia Bergqvist, Cesare Baronio, Chenge Li, Christian Brown, Christoph Loderer, Ghada Nouairia, Elin Roos, Ellinor Haglund, Eloy Vallina, Emma Danelius, Fan Yang, Faraz Vosough – thank you for all your help and wise advices, Fatemeh Madani – thank you for your help with lipid vesicles preparations, Henry Ampah-Korsah, Hongzhi Wang, Huabing Wang, Huixin Yu, Inna Rozman Grinberg, Jakob Dogan, James Cumming, Jan Aaseth, Jean-Philippe Demers, Jinming Wu, Joan Patrick – thank you for always being so friendly, smart, graceful, and for English language advices, Joanna Lachowicz, Jobst Liebau, Johannes Björnerås, Lisa Lang, Magdalena Rzepka, Margareta Sahlin, Matthiew Fielden, Maurizio Baldassarre, Médoune Sarr, Mikael Oliveberg, Mingming Xu, Monica Nordberg, Nadejda Eremina, Nicklas Österlund, Ornella Bimai, Pascal Meier, Paul Sumar, Philip Williamson, Pia Harryson, Pontus Petterson, Riccardo Diamanti, Rolf Loch, Rosita Cappai, Sabeen Survery, Sarah Leeb, Seongil Choi, Scott Ayton, Sofia Gama, Stefan Nordlund, Syed Razaul Haq, Sylwia Król, Therese Grönlund – thank you for your endless enthusiasm, for your honesty and good advices, Thibault Viennet, Tim Hofer, Vladana Vukojevic, Weihua Ye, Xin Mu – thank you all very much. If I have forgotten anyone I apologize. I have had the fortune to meet you and I will forever be grateful. Best wishes to you all. I am grateful for my former neighbors Harold Shapiro and Max Tegmark, and the cousin of my father, Stefan Rosén, for being a substantial inspiration for me by their own academic pursuits. I wish to acknowledge present and absent Family, Friends, Relatives, Mormor , Mamma, Pappa, sister Lotta, my dear husband Martin and our highly beloved cat companion Helga Tassimo Onatopp Mörman for your endless support and great love – Mahalo, you all mean the most to me. You are Everything. Aloha.

58 10. References

1. Hippius, H., and Neundörfer, G. (2003) The discovery of Alzheimer’s disease. Dialogues Clin. Neurosci. 5, 101–108 2. Selkoe, D. J., and Hardy, J. (2016) The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol. Med. 8, 595–608 3. Panza, F., Seripa, D., Solfrizzi, V., Imbimbo, B. P., Lozupone, M., Leo, A., Sardone, R., Gagliardi, G., Lofano, L., Creanza, B. C., Bisceglia, P., Daniele, A., Bellomo, A., Greco, A., and Logroscino, G. (2016) Emerging drugs to reduce abnormal β-amyloid protein in Alzheimer’s disease patients. Expert Opin. Emerg. Drugs. 21, 377–391 4. Crespi, G. A. N., Hermans, S. J., Parker, M. W., and Miles, L. A. (2015) Molecular basis for mid-region amyloid- β capture by leading Alzheimer’s disease immunotherapies. Sci. Rep. 5, 9649 5. Liu, J., Yang, B., Ke, J., Li, W., and Suen, W. C. (2016) Antibody-Based Drugs and Approaches Against Amyloid-β Species for Alzheimer’s Disease Immunotherapy. Drugs and Aging. 33, 685–697 6. Solomon, B. (2007) Antibody-mediated immunotherapy for Alzheimer’s disease. Curr. Opin. Investig. Drugs. 8, 519–524 7. Sevigny, J., Chiao, P., Bussière, T., Weinreb, P. H., Williams, L., Maier, M., Dunstan, R., Salloway, S., Chen, T., Ling, Y., Gorman, J. O., Qian, F., Arastu, M., Li, M., Chollate, S., Brennan, M. S., Quintero-monzon, O., Scannevin, R. H., Arnold, H. M., Engber, T., Rhodes, K., Ferrero, J., Hang, Y., Mikulskis, A., Grimm, J., Hock, C., Nitsch, R. M., and Sandrock, A. (2016) The antibody aducanumab reduces Aβ plaques in Alzheimer’s disease. Nat. Publ. Gr. 537, 50–56 8. Ankarcrona, M., Winblad, B., Monteiro, C., Fearns, C., Powers, E. T., Johansson, J., Westermark, G. T., Presto, J., Ericzon, B. G., and Kelly, J. W. (2016) Current and future treatment of amyloid diseases. J. Intern. Med. 280, 177–202 9. Kuznetsova, I. M., Turoverov, K. K., and Uversky, V. N. (2014) What macromolecular crowding can do to a protein. Int. J. Mol. Sci. 15, 23090–23140 10. Dobson, C. M. (2004) Principles of protein folding, misfolding and aggregation. Semin. Cell Dev. Biol. 15, 3– 16 11. Delgadillo, R. F., Mueser, T. C., Zaleta-Rivera, K., Carnes, K. A., González-Valdez, J., and Parkhurst, L. J. (2019) Detailed characterization of the solution kinetics and thermodynamics of biotin, biocytin and HABA binding to avidin and streptavidin. PLoS One. 14, e0204194 12. Schreiber, G., Haran, G., and Zhou, H. X. (2009) Fundamental aspects of Protein - Protein association kinetics. Chem. Rev. 109, 839–860 13. Gsponer, J., and Babu, M. M. (2012) Cellular Strategies for Regulating Functional and Nonfunctional Protein Aggregation. Cell Rep. 2, 1425–1437 14. Fox, L. M., and Yamamoto, A. (2015) Macroautophagy of Aggregation-Prone Proteins in Neurodegenerative Disease. in Autophagy: Cancer, Other Pathologies, Inflammation, Immunity, Infection, and Aging, pp. 117– 137, 7, 117–137 15. Babu, M. M. (2016) The contribution of intrinsically disordered regions to protein function, cellular complexity, and human disease. Biochem. Soc. Trans. 44, 1185–1200 16. Johnson, S. M., Connelly, S., Fearns, C., Powers, E. T., and Kelly, J. W. (2012) The transthyretin amyloidoses: From delineating the molecular mechanism of aggregation linked to pathology to a regulatory-agency-approved drug. J. Mol. Biol. 421, 185–203

59 17. Chiti, F., and Dobson, C. M. (2017) Protein Misfolding, Amyloid Formation, and Human Disease: A Summary of Progress Over the Last Decade. Annu. Rev. Biochem. 86, 27–68 18. Knowles, T., Vendruscolo, M., and Dobson, C. (2014) The amyloid state and its association with protein misfolding diseases. Nat Rev Mol Cell Biol. 16, 384–396 19. Tipping, K. W., van Oosten-Hawle, P., Hewitt, E. W., and Radford, S. E. (2015) Amyloid Fibres: Inert End- Stage Aggregates or Key Players in Disease? Trends Biochem. Sci. 40, 719–727 20. Hardy, J. A., and Higgins, G. A. (1992) Alzheimer’s disease: the amyloid cascade hypothesis. Science. 256, 184–185 21. Landreh, M., Sawaya, M. R., Hipp, M. S., Eisenberg, D. S., Wüthrich, K., and Hartl, F. U. (2016) The formation, function and regulation of amyloids: insights from structural biology. J. Intern. Med. 280, 164–176 22. Dobson, C. M. (2017) The amyloid phenomenon and its links with human disease. Cold Spring Harb. Perspect. Biol. 9, a023648 23. Oliveberg, M., and Shakhnovich, E. I. (2006) Folding and binding: The conformational repertoire of proteins: Folding, aggregation and structural recognition. Curr. Opin. Struct. Biol. 16, 68–70 24. Serra-Batiste, M., Ninot-Pedrosa, M., Bayoumi, M., Gairí, M., Maglia, G., and Carulla, N. (2016) Aβ42 assembles into specific β-barrel pore-forming oligomers in membrane-mimicking environments. Proc. Natl. Acad. Sci. U. S. A. 113, 10866–10871 25. Ciudad, S., Puig, E., Botzanowski, T., Meigooni, M., Arango, A. S., Do, J., Mayzel, M., Bayoumi, M., Chaignepain, S., Maglia, G., Cianferani, S., Orekhov, V., Tajkhorshid, E., Bardiaux, B., and Carulla, N. (2019) Aβ(1-42) tetramer and octamer structures reveal edge pores as a mechanism for membrane damage. bioRxiv. 10.1101/759472 26. Österlund, N., Moons, R., Ilag, L. L., Sobott, F., and Graslund, A. (2019) Native ion mobility-mass spectrometry reveals the formation of β‑barrel shaped amyloid‑β hexamers in a membrane-mimicking environment. J. Am. Chem. Soc. 141, 10440–10450 27. Dubnovitsky, A., Sandberg, A., Rahman, M. M., Benilova, I., Lendel, C., and Härd, T. (2013) Amyloid-β Protofibrils: Size, Morphology and Synaptotoxicity of an Engineered Mimic. PLoS One. 8, e66101 28. Fleming Outeiro, T., and Tetzlaff, J. (2007) Mechanisms of Disease II: Cellular Protein Quality Control. Semin. Pediatr. Neurol. 14, 15–25 29. Schopf, F. H., Biebl, M. M., and Buchner, J. (2017) The HSP90 chaperone machinery. Nat. Rev. Mol. Cell Biol. 18, 345–360 30. Chen, G., Abelein, A., Nilsson, H. E., Leppert, A., Andrade-Talavera, Y., Tambaro, S., Hemmingsson, L., Roshan, F., Landreh, M., Biverstål, H., Koeck, P. J. B., Presto, J., Hebert, H., Fisahn, A., and Johansson, J. (2017) Bri2 BRICHOS client specificity and chaperone activity are governed by assembly state. Nat. Commun. 8, 2081 31. Cohen, S. I. A., Arosio, P., Presto, J., Kurudenkandy, F. R., Biverstål, H., Dolfe, L., Dunning, C., Yang, X., Frohm, B., Vendruscolo, M., Johansson, J., Dobson, C. M., Fisahn, A., Knowles, T. P. J., and Linse, S. (2015) A molecular chaperone breaks the catalytic cycle that generates toxic Aβ oligomers. Nat. Struct. Mol. Biol. 22, 207–213 32. Leppert, A., Chen, G., and Johansson, J. (2019) BRICHOS: a chaperone with different activities depending on quaternary structure and cellular location? Amyloid. 26, 152–153 33. Nelson, A. R., Sagare, A. P., and Zlokovic, B. V. (2017) Role of clusterin in the brain vascular clearance of amyloid-β. Proc. Natl. Acad. Sci. U. S. A. 114, 8681–8682 34. Scheidt, T., Łapińska, U., Kumita, J. R., Whiten, D. R., Klenerman, D., Wilson, M. R., Cohen, S. I. A., Linse, S., Vendruscolo, M., Dobson, C. M., Knowles, T. P. J., and Arosio, P. (2019) Secondary nucleation and elongation occur at different sites on Alzheimer’s amyloid-β aggregates. Sci. Adv. 5, eaau3112 35. Månsson, C., Van Cruchten, R. T. P., Weininger, U., Yang, X., Cukalevski, R., Arosio, P., Dobson, C. M., Knowles, T., Akke, M., Linse, S., and Emanuelsson, C. (2018) Conserved S/T Residues of the Human Chaperone DNAJB6 Are Required for Effective Inhibition of Aβ42 Amyloid Fibril Formation. Biochemistry. 57, 4891–4892 36. Soto, C., and Pritzkow, S. (2018) Protein misfolding, aggregation, and conformational strains in neurodegenerative diseases. Nat. Neurosci. 21, 1332–1340 37. Lashuel, H. A., Hartley, D., Petre, B. M., Walz, T., and Jr, Lansbury, P. T. (2002) Neurodegenerative disease:

60 Amyloid pores from pathogenic mutations. Nature. 418, 291–291 38. Olzscha, H., Schermann, S. M., Woerner, A. C., Pinkert, S., Hecht, M. H., Tartaglia, G. G., Vendruscolo, M., Hayer-Hartl, M., Hartl, F. U., and Vabulas, R. M. (2011) Amyloid-like aggregates sequester numerous metastable proteins with essential cellular functions. Cell. 144, 67–78 39. Necula, M., Kayed, R., Milton, S., and Glabe, C. G. (2007) Small molecule inhibitors of aggregation indicate that amyloid β oligomerization and fibrillization pathways are independent and distinct. J. Biol. Chem. 282, 10311–10324 40. Stewart, K. L., and Radford, S. E. (2017) Amyloid plaques beyond Aβ: a survey of the diverse modulators of amyloid aggregation. Biophys. Rev. 9, 405–419 41. Gallardo, R., Ramakers, M., De Smet, F., Claes, F., Khodaparast, L., Khodaparast, L., Couceiro, J. R., Langenberg, T., Siemons, M., Nystrom, S., Young, L. J., Laine, R. F., Young, L., Radaelli, E., Benilova, I., Kumar, M., Staes, A., Desager, M., Beerens, M., Vandervoort, P., Luttun, A., Gevaert, K., Bormans, G., Dewerchin, M., Van Eldere, J., Carmeliet, P., Vande Velde, G., Verfaillie, C., Kaminski, C. F., De Strooper, B., Hammarström, P., Nilsson, K. P. R., Serpell, L., Schymkowitz, J., and Rousseau, F. (2016) De novo design of a biologically active amyloid. Science. 354, aah4949 42. Mayeux, R., and Stern, Y. (2012) Epidemiology of Alzheimer disease. Cold Spring Harb. Perspect. Med. 2, a006239 43. Guerchet, M., Prina, M., and Prince, M. (2013) Policy Brief for Heads of Government: The Global Impact of Dementia 2013–2050. Policy Br. Heads Gov. Glob. Impact Dement. 2013–2050 Publ. by Alzheimer’s Dis. Int. (ADI), London. December 2013 44. A Armstrong, R. (2019) Risk factors for Alzheimer’s disease. Folia Neuropathol. 57, 87–105 45. Tari, A. R., Norevik, C. S., Scrimgeour, N. R., Kobro-Flatmoen, A., Storm-Mathisen, J., Bergersen, L. H., Wrann, C. D., Selbæk, G., Kivipelto, M., Moreira, J. B. N., and Wisløff, U. (2019) Are the neuroprotective effects of exercise training systemically mediated? Prog. Cardiovasc. Dis. 62, 94–101 46. Kivipelto, M., Mangialasche, F., and Ngandu, T. (2018) Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat. Rev. Neurol. 14, 653–666 47. Bertram, L., and Tanzi, R. E. (2008) Thirty years of Alzheimer’s disease genetics: The implications of systematic meta-analyses. Nat. Rev. Neurosci. 9, 768–778 48. Weggen, S., and Beher, D. (2012) Molecular consequences of amyloid precursor protein and presenilin mutations causing autosomal-dominant Alzheimer’s disease. Alzheimer’s Res. Ther. 4, 9 49. Verghese, P. B., Castellano, J. M., and Holtzman, D. M. (2011) Apolipoprotein E in Alzheimer’s disease and other neurological disorders. Lancet Neurol. 10, 241–252 50. Suzuki, K., Hirakawa, A., Ihara, R., Iwata, A., Ishii, K., Ikeuchi, T., Sun, C., Donohue, M., and Iwatsubo, T. (2020) Effect of apolipoprotein E ε4 allele on the progression of cognitive decline in the early stage of Alzheimer’s disease. Alzheimer’s Dement. Transl. Res. Clin. Interv. 6, e12007 51. Jonsson, T., Atwal, J. K., Steinberg, S., Snaedal, J., Jonsson, P. V., Bjornsson, S., Stefansson, H., Sulem, P., Gudbjartsson, D., Maloney, J., Hoyte, K., Gustafson, A., Liu, Y., Lu, Y., Bhangale, T., Graham, R. R., Huttenlocher, J., Bjornsdottir, G., Andreassen, O. A., Jonsson, E. G., Palotie, A., Behrens, T. W., Magnusson, O. T., Kong, A., Thorsteinsdottir, U., Watts, R. J., and Stefansson, K. (2012) A mutation in APP protects against Alzheimer‘s disease and age-related cognitive decline. Nature. 488, 96 52. Colombo, L., Gamba, A., Cantù, L., Salmona, M., Tagliavini, F., Rondelli, V., Del Favero, E., and Brocca, P. (2017) Pathogenic Aβ A2V versus protective Aβ A2T mutation: Early stage aggregation and membrane interaction. Biophys. Chem. 229, 11–18 53. Nasica-Labouze, J., Nguyen, P., Sterpone, F., Berthoumieu, O., Buchete, N., Coté, S., De Simone, A., Doig, A., Faller, P., Garcia, A., Laio, A., Li, M., Melchionna, S., Mousseau, N., Mu, Y., Paravastu, A., Pasquali, S., Rosenman, D., Strodel, B., Tarus, B., Viles, J., Zhang, T., Wang, C., and Derreumaux, P. (2015) Amyloid β Protein and Alzheimer’s Disease: When Computer Simulations Complement Experimental Studies. Chem Rev. 115, 3518–63 54. Ballard, C., Gauthier, S., Corbett, A., Brayne, C., Aarsland, D., and Jones, E. (2011) Alzheimer’s disease. Lancet. 377, 1019–1031 55. McShane, R., Westby, M. J., Roberts, E., Minakaran, N., Schneider, L., Farrimond, L. E., Maayan, N., Ware, J.,

61 and Debarros, J. (2019) Memantine for dementia. Cochrane Database Syst. Rev. 2019, 1–446 56. Kepp, K. P. (2012) Bioinorganic Chemistry of Alzheimer’s Disease. Chem. Rev. 112, 5193–5239 57. Jagust, W. (2018) Imaging the evolution and pathophysiology of Alzheimer disease. Nat. Rev. Neurosci. 19, 687–700 58. Braak, H., and Braak, E. (1995) Staging of alzheimer’s disease-related neurofibrillary changes. Neurobiol. Aging. 16, 271–278 59. Eisele, Y. S., and Duyckaerts, C. (2016) Propagation of Aß pathology: hypotheses, discoveries, and yet unresolved questions from experimental and human brain studies. Acta Neuropathol. 131, 5–25 60. Dhiman, K., Blennow, K., Zetterberg, H., Martins, R. N., and Gupta, V. B. (2019) Cerebrospinal fluid biomarkers for understanding multiple aspects of Alzheimer’s disease pathogenesis. Cell. Mol. Life Sci. 76, 1833–1863 61. Jan, A., Gokce, O., Luthi-Carter, R., and Lashuel, H. A. (2008) The ratio of monomeric to aggregated forms of Aβ40 and Aβ42 is an important determinant of amyloid-β aggregation, fibrillogenesis, and toxicity. J. Biol. Chem. 283, 28176–28189 62. Villemagne, V. L., Doré, V., Burnham, S. C., Masters, C. L., and Rowe, C. C. (2018) Imaging tau and amyloid- β proteinopathies in Alzheimer disease and other conditions. Nat. Rev. Neurol. 14, 225–236 63. Sperling, R. A., Jack, C. R., and Aisen, P. S. (2011) Testing the right target and right drug at the right stage. Sci. Transl. Med. 3, 111cm33 64. Colom-Cadena, M., Spires-Jones, T., Zetterberg, H., Blennow, K., Caggiano, A., DeKosky, S. T., Fillit, H., Harrison, J. E., Schneider, L. S., Scheltens, P., de Haan, W., Grundman, M., van Dyck, C. H., Izzo, N. J., and Catalano, S. M. (2020) The clinical promise of biomarkers of synapse damage or loss in Alzheimer’s disease. Alzheimers. Res. Ther. 12, 21 65. Simrén, J., Ashton, N. J., Blennow, K., and Zetterberg, H. (2020) An update on fluid biomarkers for neurodegenerative diseases: recent success and challenges ahead. Curr. Opin. Neurobiol. 61, 29–39 66. Janelidze, S., Mattsson, N., Palmqvist, S., Smith, R., Beach, T. G., Serrano, G. E., Chai, X., Proctor, N. K., Eichenlaub, U., Zetterberg, H., Blennow, K., Reiman, E. M., Stomrud, E., Dage, J. L., and Hansson, O. (2020) Plasma P-tau181 in Alzheimer’s disease: relationship to other biomarkers, differential diagnosis, neuropathology and longitudinal progression to Alzheimer’s dementia. Nat. Med. 26, 379–386 67. Tiiman, A., Jelić, V., Jarvet, J., Järemo, P., Bogdanović, N., Rigler, R., Terenius, L., Gräslund, A., and Vukojević, V. (2019) Amyloidogenic Nanoplaques in Blood Serum of Patients with Alzheimer’s Disease Revealed by Time-Resolved Thioflavin T Fluorescence Intensity Fluctuation Analysis. J. Alzheimer’s Dis. 68, 571–582 68. De Strooper, B., and Karran, E. (2016) The Cellular Phase of Alzheimer’s Disease. Cell. 164, 603–615 69. Nesse, R., Finch, C., and Nunn, C. (2017) Does selection for short sleep duration explain human vulnerability to Alzheimer’s disease? Evol Med Public Heal. 2017, 39–46 70. Pawlowski, M., Meuth, S. G., and Duning, T. (2017) Cerebrospinal Fluid Biomarkers in Alzheimer’s Disease- From Brain Starch to Bench and Bedside. Diagnostics (Basel, Switzerland). 7, E42 71. Kashyap, G., Bapat, D., Das, D., Gowaikar, R., Amritkar, R. E., Rangarajan, G., Ravindranath, V., and Ambika, G. (2019) Synapse loss and progress of Alzheimer’s disease -A network model. Sci. Rep. 9, 6555 72. Michaels, T. C. T., Šarić, A., Curk, S., Bernfur, K., Arosio, P., Meisl, G., Dear, A. J., Cohen, S. I. A., Dobson, Christopher M.Vendruscolo, M., Linse, S., and Knowles, T. P. J. (2020) Dynamics of oligomer populations formed during the aggregation of Alzheimer’s Aβ42 peptide. Nat. Chem. 12, 445–451 73. Medina, M., Hernández, F., and Avila, J. (2016) New Features about Tau Function and Dysfunction. Biomolecules. 6, 21 74. Liu, P.-P., Xie, Y., Meng, X.-Y., and Kang, J.-S. (2019) History and progress of hypotheses and clinical trials for Alzheimer’s disease. Signal Transduct. Target. Ther. 4, 29 75. Panza, F., Lozupone, M., Solfrizzi, V., Watling, M., and Imbimbo, B. P. (2019) Time to test antibacterial therapy in Alzheimer’s disease. Brain. 142, 2905–2929 76. Ittner, L. M., and Götz, J. (2011) Amyloid-β and tau - A toxic pas de deux in Alzheimer’s disease. Nat. Rev. Neurosci. 12, 67–72 77. Iqbal, K., Liu, F., and Gong, C. X. (2016) Tau and neurodegenerative disease: The story so far. Nat. Rev. Neurol. 12, 15–27

62 78. Goedert, M. (1996) Tau Protein and the Neurofibrillary Pathology of Alzheimer’s Disease. in Apolipoprotein E and Alzheimer’s Disease, pp. 103–125 79. Ayton, S., Lei, P., and Bush, A. I. (2013) Metallostasis in Alzheimer’s disease. Free Radic. Biol. Med. 62, 76– 89 80. Ayton, S., Lei, P., and Bush, A. I. (2015) Biometals and Their Therapeutic Implications in Alzheimer’s Disease. Neurotherapeutics. 12, 109–120 81. Edwards, F. A. (2019) A Unifying Hypothesis for Alzheimer’s Disease: From Plaques to Neurodegeneration. Trends Neurosci. 42, 310–322 82. Sakono, M., and Zako, T. (2010) Amyloid oligomers: formation and toxicity of Abeta oligomers. FEBS J. 277, 1348–58 83. Oddo, S., Caccamo, A., Smith, I. F., Green, K. N., and LaFerla, F. M. (2006) A Dynamic Relationship between Intracellular and Extracellular Pools of Aβ. Am. J. Pathol. 168, 184–194 84. Clavaguera, F., Hench, J., Goedert, M., and Tolnay, M. (2015) Invited review: Prion-like transmission and spreading of tau pathology. Neuropathol. Appl. Neurobiol. 41, 47–58 85. Scholz, T., and Mandelkow, E. (2014) Transport and diffusion of Tau protein in neurons. Cell. Mol. Life Sci. 71, 3139–3150 86. Glenner, G. G., and Wong, C. W. (1984) Alzheimer’s disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun. 120, 885–890 87. Yiannopoulou, K. G., Anastasiou, A. I., Zachariou, V., and Pelidou, S. H. (2019) Reasons for failed trials of disease-modifying treatments for alzheimer disease and their contribution in recent research. Biomedicines. 7, E97 88. Chen, G. F., Xu, T. H., Yan, Y., Zhou, Y. R., Jiang, Y., Melcher, K., and Xu, H. E. (2017) Amyloid beta: Structure, biology and structure-based therapeutic development. Acta Pharmacol. Sin. 38, 1205–1235 89. Roche, J., Shen, Y., Lee, J. H., Ying, J., and Bax, A. (2016) Monomeric Aβ1-40 and Aβ1-42 Peptides in Solution Adopt Very Similar Ramachandran Map Distributions That Closely Resemble Random Coil. Biochemistry. 55, 762–775 90. Danielsson, J., Jarvet, J., Damberg, P., and Gräslund, A. (2005) The Alzheimer β-peptide shows temperature- dependent transitions between left-handed 31-helix, β-strand and random coil secondary structures. FEBS J. 272, 3938–3949 91. Jarvet, J., Danielsson, J., Damberg, P., Oleszczuk, M., and Gräslund, A. (2007) Positioning of the Alzheimer Aβ(1-40) peptide in SDS micelles using NMR and paramagnetic probes. J. Biomol. NMR. 39, 63–72 92. Tharp, W. G., and Sarkar, I. N. (2013) Origins of amyloid-β. BMC Genomics. 14, 290 93. Müller, U. C., Deller, T., and Korte, M. (2017) Not just amyloid: Physiological functions of the amyloid precursor protein family. Nat. Rev. Neurosci. 18, 281–298 94. C. Crdenas-Aguayo, M. del, C. Silva-Lucero, M. del, Cortes-Ortiz, M., Jimnez-Ramos, B., Gmez-Virgilio, L., Ramrez-Rodrguez, G., Vera- Arroyo, E., Fiorentino-Prez, R., Garca, U., Luna-Muoz, J., and A. Meraz Ros, M. (2014) Physiological Role of Amyloid Beta in Neural Cells: The Cellular Trophic Activity. in Neurochemistry, 10.5772/57398 95. Dawkins, E., and Small, D. H. (2014) Insights into the physiological function of the β-amyloid precursor protein: Beyond Alzheimer’s disease. J. Neurochem. 129, 756–769 96. Pearson, H. A., and Peers, C. (2006) Physiological roles for amyloid β peptides. J. Physiol. 575, 5–10 97. Sosa, L. J., Cáceres, A., Dupraz, S., Oksdath, M., Quiroga, S., and Lorenzo, A. (2017) The physiological role of the amyloid precursor protein as an adhesion molecule in the developing nervous system. J. Neurochem. 143, 11–29 98. Wild, K., August, A., Pietrzik, C. U., and Kins, S. (2017) Structure and synaptic function of metal binding to the amyloid precursor protein and its proteolytic fragments. Front. Mol. Neurosci. 10, 21 99. Wallin, C., Luo, J., Jarvet, J., Wärmländer, S. K. T. S., and Gräslund, A. (2017) The Amyloid-β Peptide in Amyloid Formation Processes: Interactions with Blood Proteins and Naturally Occurring Metal Ions. Isr J Chem. 57, 674–685 100. Frederickson, C. J., and Bush, A. I. (2001) Synaptically released zinc: Physiological functions and pathological effects. BioMetals. 14, 353–366

63 101.Lin, R., Chen, X., Li, W., Han, Y., Liu, P., and Pi, R. (2008) Exposure to metal ions regulates mRNA levels of APP and BACE1 in PC12 cells: Blockage by curcumin. Neurosci. Lett. 440, 344–347 102. Bellingham, S. A., Lahiri, D. K., Maloney, B., La Fontaine, S., Multhaup, G., and Camakaris, J. (2004) Copper Depletion Down-regulates Expression of the Alzheimer’s Disease Amyloid-β Precursor Protein Gene. J. Biol. Chem. 279, 20378–20386 103. Cater, M. a, McInnes, K. T., Li, Q.-X., Volitakis, I., La Fontaine, S., Mercer, J. F. B., and Bush, A. I. (2008) Intracellular copper deficiency increases amyloid-beta secretion by diverse mechanisms. Biochem. J. 412, 141– 152 104. Maynard, C. J., Cappai, R., Volitakis, I., Cherny, R. A., White, A. R., Beyreuther, K., Masters, C. L., Bush, A. I., and Li, Q. X. (2002) Overexpression of Alzheimer’s disease amyloid-β opposes the age-dependent elevations of brain copper and iron. J. Biol. Chem. 277, 44670–44676 105. Morley, J., Farr, S., Banks, W., Johnson, S., Yamada, K., and Xu, L. (2010) A Physiological Role for Amyloid Beta Protein: Enhancement of Learning and Memory. J Alzheimers Dis. 19, 441–9 106. Wärmländer, S., Tiiman, A., Abelein, A., Luo, J., Jarvet, J., Söderberg, K. L., Danielsson, J., and Gräslund, A. (2013) Biophysical studies of the amyloid β-peptide: Interactions with metal ions and small molecules. ChemBioChem. 14, 1692–1704 107. Wang, H., Muiznieks, L. D., Ghosh, P., Williams, D., Solarski, M., Fang, A., Ruiz-Riquelme, A., Pomès, R., Watts, J. C., Chakrabartty, A., Wille, H., Sharpe, S., and Schmitt-Ulms, G. (2017) Somatostatin binds to the human amyloid β peptide and favors the formation of distinct oligomers. Elife. 6, e28401 108. Sengupta, P., Garai, K., Sahoo, B., Shi, Y., Callaway, D. J. E., and Maiti, S. (2003) The amyloid β peptide (Aβ1- 40) is thermodynamically soluble at physiological concentrations. Biochemistry. 42, 10506–10513 109. Gremer, L., Schölzel, D., Schenk, C., Reinartz, E., Labahn, J., Ravelli, R. B. G., Tusche, M., Lopez-Iglesias, C., Hoyer, W., Heise, H., Willbold, D., and Schröder, G. F. (2017) Fibril structure of amyloid-β(1–42) by cryo– electron microscopy. Science. 358, 116–119 110. Söldner, C. A., Sticht, H., and Horn, A. H. C. (2017) Role of the N-terminus for the stability of an amyloid-β fibril with three-fold symmetry. PLoS One. 12, e0186347 111. Cheignon, C., Jones, M., Atrián-Blasco, E., Kieffer, I., Faller, P., Collin, F., and Hureau, C. (2017) Identification of key structural features of the elusive Cu-Aβ complex that generates ROS in Alzheimer’s disease. Chem. Sci. 8, 5107–5118 112. Sipe, J. D., and Cohen, A. S. (2000) Review: History of the amyloid fibril. J. Struct. Biol. 130, 88–98 113. Fowler, D. M., Koulov, A. V., Balch, W. E., and Kelly, J. W. (2007) Functional amyloid - from bacteria to humans. Trends Biochem. Sci. 32, 217–224 114. Otzen, D., and Riek, R. (2019) Functional amyloids. Cold Spring Harb. Perspect. Biol. 11, a033860 115. Marshall, K. E., and Serpell, L. C. (2009) Structural integrity of β-sheet assembly. Biochemical Society Transactions. 37, 671–676 116. Maji, S. K., Perrin, M. H., Sawaya, M. R., Jessberger, S., Vadodaria, K., Rissman, R. A., Singru, P. S., Nilsson, K. P. R., Simon, R., Schubert, D., Eisenberg, D., Rivier, J., Sawchenko, P., Vale, W., and Riek, R. (2009) Functional amyloids as natural storage of peptide hormones in pituitary secretory granules. Science. 325, 328– 332 117. Slotta, U., Hess, S., Spieß, K., Stromer, T., Serpell, L., and Scheibel, T. (2007) Spider silk and amyloid fibrils: A structural comparison. Macromol. Biosci. 7, 183–188 118. Arosio, P., Michaels, T. C. T., Linse, S., Månsson, C., Emanuelsson, C., Presto, J., Johansson, J., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2016) Kinetic analysis reveals the diversity of microscopic mechanisms through which molecular chaperones suppress amyloid formation. Nat. Commun. 7, 10948 119. Meisl, G., Michaels, T. C. T., Linse, S., and Knowles, T. P. J. (2018) Kinetic analysis of amyloid formation. in Methods in Molecular Biology, pp. 181–196, 1779, 181–196 120. Sunde, M., and Blake, C. C. F. (1998) From the globular to the fibrous state: protein structure and structural conversion in amyloid formation. Q. Rev. Biophys. 31, 1–39 121. Michaels, T. C. T., Šarić, A., Habchi, J., Chia, S., Meisl, G., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2018) Chemical Kinetics for Bridging Molecular Mechanisms and Macroscopic Measurements of Amyloid Fibril Formation. Annu. Rev. Phys. Chem. 69, 273–298

64 122. Meier, B. H., Riek, R., and Böckmann, A. (2017) Emerging Structural Understanding of Amyloid Fibrils by Solid-State NMR. Trends Biochem. Sci. 42, 777–787 123. Schmidt, M., Rohou, A., Lasker, K., Yadav, J., Schiene-Fischer, C., Fändrich, M., and Grigorieff, N. (2015) Peptide dimer structure in an Aβ(1–42) fibril visualized with cryo-EM. Proc Natl Acad Sci U S A. 112, 11858– 63 124. Wälti, M. A., Ravotti, F., Arai, H., Glabe, C. G., Wall, J. S., Böckmann, A., Güntert, P., Meier, B. H., and Riek, R. (2016) Atomic-resolution structure of a disease-relevant Aβ(1-42) amyloid fibril. Proc. Natl. Acad. Sci. U. S. A. 113, E4976–E4984 125. Silvers, R., Colvin, M. T., Frederick, K. K., Jacavone, A. C., Lindquist, S., Linse, S., and Griffin, R. G. (2017) Aggregation and Fibril Structure of AβM01-42 and Aβ1-42. Biochemistry. 56, 4850–4859 126. Sawaya, M. R., Sambashivan, S., Nelson, R., Ivanova, M. I., Sievers, S. A., Apostol, M. I., Thompson, M. J., Balbirnie, M., Wiltzius, J. J. W., McFarlane, H. T., Madsen, A. Ø., Riekel, C., and Eisenberg, D. (2007) Atomic structures of amyloid cross-beta spines reveal varied steric zippers. Nature. 447, 453–7 127. Fitzpatrick, A. W. P., Falcon, B., He, S., Murzin, A. G., Murshudov, G., Garringer, H. J., Crowther, R. A., Ghetti, B., Goedert, M., and Scheres, S. H. W. (2017) Cryo-EM structures of tau filaments from Alzheimer’s disease. Nature. 547, 185–190 128. Lu, J. X., Qiang, W., Yau, W. M., Schwieters, C. D., Meredith, S. C., and Tycko, R. (2013) Molecular structure of β-amyloid fibrils in Alzheimer’s disease brain tissue. Cell. 154, 1257 129. Colvin, M. T., Silvers, R., Ni, Q. Z., Can, T. V., Sergeyev, I., Rosay, M., Donovan, K. J., Michael, B., Wall, J., Linse, S., and Griffin, R. G. (2016) Atomic Resolution Structure of Monomorphic Aβ42 Amyloid Fibrils. J. Am. Chem. Soc. 138, 9663–9674 130. Meisl, G., Yang, X., Frohm, B., Knowles, T. P. J., and Linse, S. (2016) Quantitative analysis of intrinsic and extrinsic factors in the aggregation mechanism of Alzheimer-associated Aβ-peptide. Sci. Rep. 6, 18728 131. Abelein, A., Lang, L., Lendel, C., Gräslund, A., and Danielsson, J. (2012) Transient small molecule interactions kinetically modulate amyloid β peptide self-assembly. FEBS Lett. 586, 3991–3995 132. Abelein, A., Jarvet, J., Barth, A., Gräslund, A., and Danielsson, J. (2016) Ionic Strength Modulation of the Free Energy Landscape of Aβ40 Peptide Fibril Formation. J. Am. Chem. Soc. 138, 6893–6902 133. Meisl, G., Yang, X., Dobson, C. M., Linse, S., and Knowles, T. P. J. (2017) Modulation of electrostatic interactions to reveal a reaction network unifying the aggregation behaviour of the Aβ42 peptide and its variants. Chem. Sci. 8, 4352–4362 134. Huang, X., Atwood, C. S., Moir, R. D., Hartshorn, M. A., Tanzi, R. E., and Bush, A. I. (2004) Trace metal contamination initiates the apparent auto-aggregation, amyloidosis, and oligomerization of Alzheimer’s Aβ peptides. J. Biol. Inorg. Chem. 9, 954–960 135. Boeynaems, S., Alberti, S., Fawzi, N. L., Mittag, T., Polymenidou, M., Rousseau, F., Schymkowitz, J., Shorter, J., Wolozin, B., Van Den Bosch, L., Tompa, P., and Fuxreiter, M. (2018) Protein Phase Separation: A New Phase in Cell Biology. Trends Cell Biol. 28, 420–435 136. Elbaum-Garfinkle, S. (2019) Matter over mind: Liquid phase separation and neurodegeneration. J. Biol. Chem. 294, 7160–7168 137. Ambadipudi, S., Biernat, J., Riedel, D., Mandelkow, E., and Zweckstetter, M. (2017) Liquid-liquid phase separation of the microtubule-binding repeats of the Alzheimer-related protein Tau. Nat. Commun. 8, 275 138. Aleksis, R., Oleskovs, F., Jaudzems, K., Pahnke, J., and Biverstål, H. (2017) Structural studies of amyloid-β peptides: Unlocking the mechanism of aggregation and the associated toxicity. Biochimie. 140, 176–192 139. Eisenberg, D., and Jucker, M. (2012) The amyloid state of proteins in human diseases. Cell. 148, 1188–1203 140. Morris, K. L., and Serpell, L. C. (2012) X-ray fibre diffraction studies of amyloid fibrils. Methods Mol. Biol. 849, 121–135 141. Hasecke, F., Miti, T., Perez, C., Barton, J., Schölzel, D., Gremer, L., Grüning, C. S. R., Matthews, G., Meisl, G., Knowles, T. P. J., Willbold, D., Neudecker, P., Heise, H., Ullah, G., Hoyer, W., and Muschol, M. (2018) Origin of metastable oligomers and their effects on amyloid fibril self-assembly. Chem. Sci. 9, 5937–5948 142. Spehar, K., Ding, T., Sun, Y., Kedia, N., Lu, J., Nahass, G. R., Lew, M. D., and Bieschke, J. (2018) Super- resolution Imaging of Amyloid Structures over Extended Times by Using Transient Binding of Single Thioflavin T Molecules. ChemBioChem. 19, 1944–1948

65 143. Tomic, J. L., Pensalfini, A., Head, E., and Glabe, C. G. (2009) Soluble fibrillar oligomer levels are elevated in Alzheimer’s disease brain and correlate with cognitive dysfunction. Neurobiol. Dis. 35, 352–358 144. Limbocker, R., Chia, S., Ruggeri, F., Perni, M., Cascella, R., Heller, G., Meisl, G., Mannini, B., Habchi, J., Michaels, T., Challa, P., Ahn, M., Casford, S., Fernando, N., Xu, C., Kloss, N., Cohen, S., Kumita, J., Cecchi, C., Zasloff, M., Linse, S., Knowles, T., Chiti, F., Vendruscolo, M., and Dobson, C. (2019) Trodusquemine enhances Aβ42 aggregation but suppresses its toxicity by displacing oligomers from cell membranes. Nat Commun. 10, 225 145. Cline, E. N., Bicca, M. A., Viola, K. L., and Klein, W. L. (2018) The Amyloid-β Oligomer Hypothesis: Beginning of the Third Decade. J. Alzheimer’s Dis. 64, S567–S610 146. Nisbet, R. M., Polanco, J. C., Ittner, L. M., and Götz, J. (2015) Tau aggregation and its interplay with amyloid- β. Acta Neuropathol. 129, 207–220 147. LaFerla, F. M., Akbari, Y., Murphy, M. P., Oddo, S., Kayed, R., Mattson, M. P., Shepherd, J. D., Golde, T. E., Caccamo, A., and Metherate, R. (2004) Triple-Transgenic Model of Alzheimer’s Disease with Plaques and Tangles. Neuron. 39, 409–421 148. Kondo, T., Asai, M., Tsukita, K., Kutoku, Y., Ohsawa, Y., Sunada, Y., Imamura, K., Egawa, N., Yahata, N., Okita, K., Takahashi, K., Asaka, I., Aoi, T., Watanabe, A., Watanabe, K., Kadoya, C., Nakano, R., Watanabe, D., Maruyama, K., Hori, O., Hibino, S., Choshi, T., Nakahata, T., Hioki, H., Kaneko, T., Naitoh, M., Yoshikawa, K., Yamawaki, S., Suzuki, S., Hata, R., Ueno, S. I., Seki, T., Kobayashi, K., Toda, T., Murakami, K., Irie, K., Klein, W. L., Mori, H., Asada, T., Takahashi, R., Iwata, N., Yamanaka, S., and Inoue, H. (2013) Modeling Alzheimer’s disease with iPSCs reveals stress phenotypes associated with intracellular Aβ and differential drug responsiveness. Cell Stem Cell. 12, 487–496 149. Penney, J., Ralvenius, W. T., and Tsai, L. H. (2020) Modeling Alzheimer’s disease with iPSC-derived brain cells. Mol. Psychiatry. 25, 148–167 150. Gonzalez, C., Armijo, E., Bravo-Alegria, J., Becerra-Calixto, A., Mays, C. E., and Soto, C. (2018) Modeling amyloid beta and tau pathology in human cerebral organoids. Mol. Psychiatry. 23, 2363–2374 151. Lapasset, L., Milhavet, O., Prieur, A., Besnard, E., Babled, A., Ät-Hamou, N., Leschik, J., Pellestor, F., Ramirez, J. M., De Vos, J., Lehmann, S., and Lemaitre, J. M. (2011) Rejuvenating senescent and centenarian human cells by reprogramming through the pluripotent state. Genes Dev. 25, 2248–2253 152. Qian, X., Song, H., and Ming, G. L. (2019) Brain organoids: Advances, applications and challenges. Development. 146, dev166074 153. Fändrich, M. (2012) Oligomeric intermediates in amyloid formation: Structure determination and mechanisms of toxicity. J. Mol. Biol. 421, 427–440 154. Kayed, R., and Lasagna-Reeves, C. A. (2013) Molecular mechanisms of amyloid oligomers toxicity. J. Alzheimer’s Dis. 33, S67–78 155. Skaper, S. D., Facci, L., Zusso, M., and Giusti, P. (2017) Synaptic Plasticity, Dementia and Alzheimer Disease. CNS Neurol. Disord. - Drug Targets. 16, 220–233 156. Rauk, A. (2008) Why is the amyloid beta peptide of Alzheimer’s disease neurotoxic? Dalton Trans. 14, 1273– 1282 157. Atamna, H., and Boyle, K. (2006) Amyloid-beta peptide binds with heme to form a peroxidase: relationship to the cytopathologies of Alzheimer’s disease. Proc. Natl. Acad. Sci. U. S. A. 103, 3381–3386 158. Brothers, H. M., Gosztyla, M. L., and Robinson, S. R. (2018) The physiological roles of amyloid-β peptide hint at new ways to treat Alzheimer’s disease. Front. Aging Neurosci. 10, 118 159. Gosztyla, M. L., Brothers, H. M., and Robinson, S. R. (2018) Alzheimer’s Amyloid-β is an Antimicrobial Peptide: A Review of the Evidence. J. Alzheimer’s Dis. 62, 1495–1506 160. Spitzer, P., Condic, M., Herrmann, M., Oberstein, T. J., Scharin-Mehlmann, M., Gilbert, D. F., Friedrich, O., Grömer, T., Kornhuber, J., Lang, R., and Maler, J. M. (2016) Amyloidogenic amyloid-β-peptide variants induce microbial agglutination and exert antimicrobial activity. Sci. Rep. 6, 32228 161. De Lorenzi, E., Chiari, M., Colombo, R., Cretich, M., Sola, L., Vanna, R., Gagni, P., Bisceglia, F., Morasso, C., Lin, J. S., Lee, M., Mcgeer, P. L., and Barron, A. E. (2017) Evidence that the human innate immune peptide LL- 37 may be a binding partner of amyloid-β and inhibitor of fibril assembly. J. Alzheimer’s Dis. 59, 1213–1226 162. Moir, R. D., Lathe, R., and Tanzi, R. E. (2018) The antimicrobial protection hypothesis of Alzheimer’s disease.

66 Alzheimer’s Dement. 14, 1602–1614 163. Al-Adwani, S., Wallin, C., Balhuizen, M. D., Veldhuizen, E. J. A., Coorens, M., Landreh, M., Végvári, Á., Smith, M. E., Qvarfordt, I., Lindén, A., Gräslund, A., Agerberth, B., and Bergman, P. (2020) Studies on citrullinated LL-37: detection in human airways, antibacterial effects and biophysical properties. Sci. Rep. 10, 2376 164. Park, S. C., Moon, J. C., Shin, S. Y., Son, H., Jung, Y. J., Kim, N. H., Kim, Y. M., Jang, M. K., and Lee, J. R. (2016) Functional characterization of alpha-synuclein protein with antimicrobial activity. Biochem. Biophys. Res. Commun. 478, 924–928 165. Ji, L., Zhao, X., Lu, W., Zhang, Q., and Hua, Z. (2016) Intracellular Aβ and its Pathological Role in Alzheimer’s Disease: Lessons from Cellular to Animal Models. Curr. Alzheimer Res. 13, 621–630 166. Kawahara, M., Kato-Negishi, M., and Tanaka, K. (2020) Amyloids: Regulators of Metal Homeostasis in the Synapse. Molecules. 25, 1441 167. Butterfield, D. A., Drake, J., Pocernich, C., and Castegna, A. (2001) Evidence of oxidative damage in Alzheimer’s disease brain: central role for amyloid beta-peptide. Trends Mol. Med. 7, 548–554 168. Butterfield, S. M., and Lashuel, H. a. (2010) Amyloidogenic protein-membrane interactions: Mechanistic insight from model systems. Angew. Chemie - Int. Ed. 49, 5628–5654 169. Lutter, L., Serpell, C. J., Tuite, M. F., and Xue, W. F. (2019) The molecular lifecycle of amyloid – Mechanism of assembly, mesoscopic organisation, polymorphism, suprastructures, and biological consequences. Biochim. Biophys. Acta - Proteins Proteomics. 1867, 140257 170. Yoshiike, Y., Tanemura, K., Murayama, O., Akagi, T., Murayama, M., Sato, S., Sun, X., Tanaka, N., and Takashima, A. (2001) New Insights on How Metals Disrupt Amyloid β-Aggregation and Their Effects on Amyloid-β Cytotoxicity. J. Biol. Chem. 276, 32293–32299 171. Vivoli Vega, M., Cascella, R., Chen, S. W., Fusco, G., De Simone, A., Dobson, C. M., Cecchi, C., and Chiti, F. (2019) The Toxicity of Misfolded Protein Oligomers Is Independent of Their Secondary Structure. ACS Chem. Biol. 14, 1593–1600 172. Reiss, A. B., Arain, H. A., Stecker, M. M., Siegart, N. M., and Kasselman, L. J. (2018) Amyloid toxicity in Alzheimer’s disease. Rev. Neurosci. 29, 613–627 173. Atwood, C. S., Obrenovich, M. E., Liu, T., Chan, H., Perry, G., Smith, M. A., and Martins, R. N. (2003) Amyloid-β: a chameleon walking in two worlds: a review of the trophic and toxic properties of amyloid-β. Brain Res. Rev. 43, 1–16 174. Marshall, K. E., Marchante, R., Xue, W. F., and Serpell, L. C. (2014) The relationship between amyloid structure and cytotoxicity. Prion. 8, 192–196 175. Quist, A., Doudevski, I., Lin, H., Azimova, R., Ng, D., Frangione, B., Kagan, B., Ghiso, J., and Lal, R. (2005) Amyloid ion channels: A common structural link for protein-misfolding disease. Proc. Natl. Acad. Sci. U. S. A. 102, 10427–10432 176. Sciacca, M. F. M., Kotler, S. A., Brender, J. R., Chen, J., Lee, D. K., and Ramamoorthy, A. (2012) Two-step mechanism of membrane disruption by Aβ through membrane fragmentation and pore formation. Biophys. J. 103, 702–710 177. Cheignon, C., Tomas, M., Bonnefont-Rousselot, D., Faller, P., Hureau, C., and Collin, F. (2018) Oxidative stress and the amyloid beta peptide in Alzheimer’s disease. Redox Biol. 14, 450–464 178. Sushma, and Mondal, A. C. (2019) Role of GPCR signaling and calcium dysregulation in Alzheimer’s disease. Mol. Cell. Neurosci. 101, 103414 179. Viña, J., and Lloret, A. (2010) Why women have more Alzheimer’s disease than men: Gender and mitochondrial toxicity of amyloid-β peptide. J. Alzheimer’s Dis. 20, S527–33 180. Cardoso, S., Carvalho, C., Correia, S. C., Seiça, R. M., and Moreira, P. I. (2016) Alzheimer’s Disease: From Mitochondrial Perturbations to Mitochondrial Medicine. Brain Pathol. 26, 632–647 181. Gao, Q., Wu, G., and Lai, K. W. C. (2020) Cholesterol Modulates the Formation of the Aβ Ion Channel in Lipid Bilayers. Biochemistry. 59, 992–998 182. Banerjee, S., and Mukherjee, S. (2018) Cholesterol: A Key in the Pathogenesis of Alzheimer’s Disease. ChemMedChem. 13, 1742–1743 183. Wong, B. X., Hung, Y. H., Bush, A. I., and Duce, J. A. (2014) Metals and cholesterol: Two sides of the same

67 coin in Alzheimer’s disease pathology. Front Aging Neurosci. 6, 91 184. Castellano, J. M., Kim, J., Stewart, F. R., Jiang, H., DeMattos, R. B., Patterson, B. W., Fagan, A. M., Morris, J. C., Mawuenyega, K. G., Cruchaga, C., Goate, A. M., Bales, K. R., Paul, S. M., Bateman, R. J., and Holtzman, D. M. (2011) Human apoE isoforms differentially regulate brain amyloid-β peptide clearance. Sci. Transl. Med. 3, 89ra57 185. Robbins, J. P., Perfect, L., Ribe, E. M., Maresca, M., Dangla-Valls, A., Foster, E. M., Killick, R., Nowosiad, P., Reid, M. J., Polit, L. D., Nevado, A. J., Ebner, D., Bohlooly-Y, M., Buckley, N., Pangalos, M. N., Price, J., and Lovestone, S. (2018) Clusterin Is Required for β-Amyloid Toxicity in Human iPSC-Derived Neurons. Front. Neurosci. 12, 504 186. Jackson, R. J., Rose, J., Tulloch, J., Henstridge, C., Smith, C., and Spires-Jones, T. L. (2019) Clusterin accumulates in synapses in Alzheimer’s disease and is increased in apolipoprotein E4 carriers. Brain Commun. 1, fcz003 187. Linse, S. S., Scheidt, T., Bernfur, K., Vendruscolo, M., Dobson, C. M., Cohen, S. I. A., Sileikis, E., Lundquist, M., Qian, F., O’Malley, T. T., Bussière, T., Weinreb, P. H., Xu, C. K., Meisl, G., Devenish, S., Knowles, T. P. J., and Hansson, O. (2019) Kinetic fingerprint of antibody therapies predicts outcomes of Alzheimer clinical trials. bioRxiv. 10.1101/815308 188. Shea, D., Hsu, C. C., Bi, T. M., Paranjapye, N., Childers, M. C., Cochran, J., Tomberlin, C. P., Wang, L., Paris, D., Zonderman, J., Varani, G., Link, C. D., Mullan, M., and Daggett, V. (2019) α-Sheet secondary structure in amyloid β-peptide drives aggregation and toxicity in Alzheimer’s disease. Proc. Natl. Acad. Sci. U. S. A. 116, 8895–8900 189. Dominy, S. S., Lynch, C., Ermini, F., Benedyk, M., Marczyk, A., Konradi, A., Nguyen, M., Haditsch, U., Raha, D., Griffin, C., Holsinger, L. J., Arastu-Kapur, S., Kaba, S., Lee, A., Ryder, M. I., Potempa, B., Mydel, P., Hellvard, A., Adamowicz, K., Hasturk, H., Walker, G. D., Reynolds, E. C., Faull, R. L. M., Curtis, M. A., Dragunow, M., and Potempa, J. (2019) Porphyromonas gingivalis in Alzheimer’s disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Sci. Adv. 5, eaau3333 190. Meisl, G., Rajah, L., Cohen, S. A. I., Pfammatter, M., Šarić, A., Hellstrand, E., Buell, A. K., Aguzzi, A., Linse, S., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2017) Scaling behaviour and rate-determining steps in filamentous self-assembly. Chem. Sci. 8, 7087–7097 191. Scheckel, C., and Aguzzi, A. (2018) Prions, prionoids and protein misfolding disorders. Nat. Rev. Genet. 19, 405–418 192. Meisl, G., Yang, X., Hellstrand, E., Frohm, B., Kirkegaard, J. B., Cohen, S. I. A., Dobson, C. M., Linse, S., and Knowles, T. P. J. (2014) Differences in nucleation behavior underlie the contrasting aggregation kinetics of the Aβ40 and Aβ42 peptides. Proc. Natl. Acad. Sci. U. S. A. 111, 9384–9389 193. Cohen, S. I. A., Linse, S., Luheshi, L. M., Hellstrand, E., White, D. A., Rajah, L., Otzen, D. E., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2013) Proliferation of amyloid-β42 aggregates occurs through a secondary nucleation mechanism. Proc. Natl. Acad. Sci. 110, 9758–9763 194. Törnquist, M., Michaels, T. C. T., Sanagavarapu, K., Yang, X., Meisl, G., Cohen, S. I. A., Knowles, T. P. J., and Linse, S. (2018) Secondary nucleation in amyloid formation. Chem. Commun. 54, 8667–8684 195. Esler, W. P., Stimson, E. R., Jennings, J. M., Vinters, H. V., Ghilardi, J. R., Lee, J. P., Mantyh, P. W., and Maggio, J. E. (2000) Alzheimer’s disease amyloid propagation by a template-dependent dock- lock mechanism. Biochemistry. 39, 6288–6295 196. Cannon, M. J., Williams, A. D., Wetzel, R., and Myszka, D. G. (2004) Kinetic analysis of beta-amyloid fibril elongation. Anal. Biochem. 328, 67–75 197. Qiang, W., Kelley, K., and Tycko, R. (2013) Polymorph-specific kinetics and thermodynamics of β-amyloid fibril growth. J. Am. Chem. Soc. 135, 6860–6871 198. Rodriguez, R. A., Chen, L. Y., Plascencia-Villa, G., and Perry, G. (2018) Thermodynamics of Amyloid-β Fibril Elongation: Atomistic Details of the Transition State. ACS Chem. Neurosci. 9, 783–789 199. Gurry, T., and Stultz, C. M. (2014) Mechanism of amyloid-β fibril elongation. Biochemistry. 53, 6981–6991 200. Cohen, S. I. A., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2012) From macroscopic measurements to microscopic mechanisms of protein aggregation. J. Mol. Biol. 421, 160–171 201. Meisl, G., Kirkegaard, J. B., Arosio, P., Michaels, T. C. T., Vendruscolo, M., Dobson, C. M., Linse, S., and

68 Knowles, T. P. J. (2016) Molecular mechanisms of protein aggregation from global fitting of kinetic models. Nat. Protoc. 11, 252–272 202. Biancalana, M., and Koide, S. (2010) Molecular mechanism of Thioflavin-T binding to amyloid fibrils. Biochim. Biophys. Acta - Proteins Proteomics. 1804, 1405–1412 203. Gade Malmos, K., Blancas-Mejia, L. M., Weber, B., Buchner, J., Ramirez-Alvarado, M., Naiki, H., and Otzen, D. (2017) ThT 101: a primer on the use of thioflavin T to investigate amyloid formation. Amyloid. 24, 1–16 204. Klingstedt, T., Åslund, A., Simon, R. A., Johansson, L. B. G., Mason, J. J., Nyström, S., Hammarström, P., and Nilsson, K. P. R. (2011) Synthesis of a library of oligothiophenes and their utilization as fluorescent ligands for spectral assignment of protein aggregates. Org. Biomol. Chem. 9, 8356–8370 205. Arosio, P., Knowles, T. P. J., and Linse, S. (2015) On the lag phase in amyloid fibril formation. Phys. Chem. Chem. Phys. 17, 7606–7618 206. Tiiman, A., Jarvet, J., Gräslund, A., and Vukojevic, V. (2015) Heterogeneity and intermediates turnover during amyloid-β (Aβ) peptide aggregation studied by Fluorescence Correlation Spectroscopy. Biochemistry. 54, 7203–7211 207. Knowles, T. P. J., Waudby, C. A., Devlin, G. L., Cohen, S. I. A., Aguzzi, A., Vendruscolo, M., Terentjev, E. M., Welland, M. E., and Dobson, C. M. (2009) An analytical solution to the kinetics of breakable filament assembly. Science. 326, 1533–1537 208. Cohen, S. I. A., Vendruscolo, M., Welland, M. E., Dobson, C. M., Terentjev, E. M., and Knowles, T. P. J. (2011) Nucleated polymerization with secondary pathways. I. Time evolution of the principal moments. J. Chem. Phys. 135, 065105 209. Arosio, P., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2014) Chemical kinetics for drug discovery to combat protein aggregation diseases. Trends Pharmacol. Sci. 35, 127–135 210. Grüning, C. S. R., Klinker, S., Wolff, M., Schneider, M., Toksöz, K., Klein, A. N., Nagel-Steger, L., Willbold, D., and Hoyer, W. (2013) The off-rate of monomers dissociating from amyloid-β protofibrils. J. Biol. Chem. 288, 37104–37111 211. Cukalevski, R., Yang, X., Meisl, G., Weininger, U., Bernfur, K., Frohm, B., Knowles, T. P. J., and Linse, S. (2015) The Aβ40 and Aβ42 peptides self-assemble into separate homomolecular fibrils in binary mixtures but cross-react during primary nucleation. Chem. Sci. 6, 4215–4233 212. Frankel, R., Törnquist, M., Meisl, G., Hansson, O., Andreasson, U., Zetterberg, H., Blennow, K., Frohm, B., Cedervall, T., Knowles, T. P. J., Leiding, T., and Linse, S. (2019) Autocatalytic amplification of Alzheimer- associated Aβ42 peptide aggregation in human cerebrospinal fluid. Commun. Biol. 2, 365 213. Yang, X., Meisl, G., Frohm, B., Thulin, E., Knowles, T. P. J., and Linse, S. (2018) On the role of sidechain size and charge in the aggregation of A β 42 with familial mutations . Proc. Natl. Acad. Sci. 115, E5849–E5858 214. Saunders, J. C., Young, L. M., Mahood, R. A., Jackson, M. P., Revill, C. H., Foster, R. J., Smith, D. A., Ashcroft, A. E., Brockwell, D. J., and Radford, S. E. (2016) An in vivo platform for identifying inhibitors of protein aggregation. Nat. Chem. Biol. 12, 94–101 215. Velander, P., Wu, L., Henderson, F., Zhang, S., Bevan, D. R., and Xu, B. (2017) Natural product-based amyloid inhibitors. Biochem. Pharmacol. 139, 40–55 216. Chainoglou, E., and Hadjipavlou-Litina, D. (2020) Curcumin in health and diseases: Alzheimer’s disease and curcumin analogues, derivatives, and hybrids. Int. J. Mol. Sci. 21, E1975 217. Smid, S. D., Maag, J. L., and Musgrave, I. F. (2012) Dietary polyphenol-derived protection against neurotoxic β-amyloid protein: From molecular to clinical. Food Funct. 3, 1242–1250 218. Faller, P., Hureau, C., and Berthoumieu, O. (2013) Role of metal ions in the self-assembly of the Alzheimer’s amyloid-β peptide. Inorg. Chem. 52, 12193–206 219. Lindberg, D. J., Wesén, E., Björkeroth, J., Rocha, S., and Esbjörner, E. K. (2017) Lipid membranes catalyse the fibril formation of the amyloid-β (1–42) peptide through lipid-fibril interactions that reinforce secondary pathways. Biochim. Biophys. Acta - Biomembr. 1859, 1921–1929 220. Tang, M., and Taghibiglou, C. (2017) The Mechanisms of Action of Curcumin in Alzheimer’s Disease. J. Alzheimer’s Dis. 58, 1003–1016 221. Sharma, A., Pachauri, V., and Flora, S. J. S. (2018) Advances in multi-functional ligands and the need for metal- related pharmacology for the management of Alzheimer disease. Front. Pharmacol. 9, 1247

69 222. Habchi, J., Arosio, P., Perni, M., Costa, A. R., Yagi-Utsumi, M., Joshi, P., Chia, S., Cohen, S. I. A., Müller, M. B. D., Linse, S., Nollen, E. A. A., Dobson, C. M., Knowles, T. P. J., and Vendruscolo, M. (2016) An anticancer drug suppresses the primary nucleation reaction that initiates the production of the toxic Ab42 aggregates linked with Alzheimer’s disease. Sci. Adv. 2, e1501244 223. Tjernberg, L. O., Näslund, J., Lindqvist, F., Johansson, J., Karlström, a R., Thyberg, J., Terenius, L., and Nordstedt, C. (1996) Arrest of beta-amyloid fibril formation by a pentapeptide ligand. J. Biol. Chem. 271, 8545– 8548 224. Ma, G., Wang, E., Wei, H., Wei, K., Zhu, P., and Liu, Y. (2013) PtCl2(phen) disrupts the metal ions binding to amyloid-β peptide. Metallomics. 5, 879–87 225. Barnham, K. J., Kenche, V. B., Ciccotosto, G. D., Smith, D. P., Tew, D. J., Liu, X., Perez, K., Cranston, G. A., Johanssen, T. J., Volitakis, I., Bush, A. I., Masters, C. L., White, A. R., Smith, J. P., Cherny, R. A., and Cappai, R. (2008) Platinum-based inhibitors of amyloid-beta as therapeutic agents for Alzheimer’s disease. Proc. Natl. Acad. Sci. U. S. A. 105, 6813–6818 226. Islam, T., Gharibyan, A. L., Golchin, S. A., Pettersson, N., Brännström, K., Hedberg, I., Virta, M. M., Olofsson, L., and Olofsson, A. (2020) Apolipoprotein E impairs amyloid-β fibril elongation and maturation. FEBS J. 287, 1208–1219 227. Luo, J., Wärmländer, S. K. T. S., Gräslund, A., and Abrahams, J. P. (2016) Cross-interactions between the Alzheimer disease amyloid-β peptide and other amyloid proteins: A further aspect of the amyloid cascade hypothesis. J. Biol. Chem. 291, 16485–16493 228. Gharibyan, A. L., Islam, T., Pettersson, N., Golchin, S. A., Lundgren, J., Johansson, G., Genot, M., Schultz, N., Wennström, M., and Olofsson, A. (2020) Apolipoprotein E interferes with IAPP aggregation and protects pericytes from IAPP-Induced Toxicity. Biomolecules. 10, E134 229. Horvath, I., Iashchishyn, I. A., Moskalenko, R. A., Wang, C., Wärmländer, S. K. T. S., Wallin, C., Gräslund, A., Kovacs, G. G., and Morozova-Roche, L. A. (2018) Co-aggregation of pro-inflammatory S100A9 with α- synuclein in Parkinson’s disease: Ex vivo and in vitro studies. J. Neuroinflammation. 15, 172 230. Ghadami, S. A., Chia, S., Ruggeri, F. S., Meisl, G., Bemporad, F., Habchi, J., Cascella, R., Dobson, C. M., Vendruscolo, M., Knowles, T. P. J., and Chiti, F. (2020) Transthyretin Inhibits Primary and Secondary Nucleations of Amyloid-β Peptide Aggregation and Reduces the Toxicity of Its Oligomers. Biomacromolecules. 21, 1112–1125 231. Chia, S., Flagmeier, P., Habchi, J., Lattanzi, V., Linse, S., Dobson, C. M., Knowles, T. P. J., and Vendruscolo, M. (2017) Monomeric and fibrillar α-synuclein exert opposite effects on the catalytic cycle that promotes the proliferation of Aβ42 aggregates. Proc. Natl. Acad. Sci. U. S. A. 114, 8005–8010 232. Long, K., Williams, T. L., and Urbanc, B. (2019) Insulin Inhibits Aβ42 Aggregation and Prevents Aβ42- Induced Membrane Disruption. Biochemistry. 58, 4519–4529 233. Kronqvist, N., Sarr, M., Lindqvist, A., Nordling, K., Otikovs, M., Venturi, L., Pioselli, B., Purhonen, P., Landreh, M., Biverstål, H., Toleikis, Z., Sjöberg, L., Robinson, C. V., Pelizzi, N., Jörnvall, H., Hebert, H., Jaudzems, K., Curstedt, T., Rising, A., and Johansson, J. (2017) Efficient protein production inspired by how spiders make silk. Nat. Commun. 8, 15504 234. Abelein, A., Chen, G., Kitoka, K., Aleksis, R., Oleskovs, F., Sarr, M., Landreh, M., Pahnke, J., Nordling, K., Kronqvist, N., Jaudzems, K., Rising, A., Johansson, J., and Biverstål, H. (2020) High-yield Production of Amyloid-β Peptide Enabled by a Customized Spider Silk Domain. Sci. Rep. 10, 235 235. Walsh, D. M., Thulin, E., Minogue, A. M., Gustavsson, N., Pang, E., Teplow, D. B., and Linse, S. (2009) A facile method for expression and purification of the Alzheimer’s disease-associated amyloid β-peptide. FEBS J. 276, 1266–1281 236. Woody, R. W. (1995) Circular Dichroism. Methods Enzymol. 246, 34–71 237. Kelly, S. M., Jess, T. J., and Price, N. C. (2005) How to study proteins by circular dichroism. Biochim. Biophys. Acta - Proteins Proteomics. 1751, 119–139 238. Johnson, W. C. (1988) Secondary structure of proteins through circular dichroism spectroscopy. Annu. Rev. Biophys. Biophys. Chem. 17, 145–166 239. Micsonai, A., Wien, F., Bulyáki, É., Kun, J., Moussong, É., Lee, Y. H., Goto, Y., Réfrégiers, M., and Kardos, J. (2018) BeStSel: A web server for accurate protein secondary structure prediction and fold recognition from the

70 circular dichroism spectra. Nucleic Acids Res. 46, W315–W322 240. Greenfield, N., and Fasman, G. D. (1969) Computed Circular Dichroism Spectra for the Evaluation of Protein Conformation. Biochemistry. 8, 4108–4116 241. Alies, B., Renaglia, E., Rózga, M., Bal, W., Faller, P., and Hureau, C. (2013) Cu(II) affinity for the Alzheimer’s peptide: Tyrosine fluorescence studies revisited. Anal. Chem. 85, 1501–1508 242. Tiiman, A., Luo, J., Wallin, C., Olsson, L., Lindgren, J., Jarvet, J., Roos, P., Sholts, S. B., Rahimipour, S., Abrahams, J. P., Karlström Eriksson, A., Gräslund, A., and Wärmländer, S. K. T. S. (2016) Specific Binding of Cu(II) Ions to Amyloid-Beta Peptides Bound to Aggregation-Inhibiting Molecules or SDS Micelles Creates Complexes that Generate Radical Oxygen Species. J. Alzheimer’s Dis. 54, 971–982 243. Xue, C., Lin, T. Y., Chang, D., and Guo, Z. (2017) Thioflavin T as an amyloid dye: Fibril quantification, optimal concentration and effect on aggregation. R. Soc. Open Sci. 4, 160696 244. Stsiapura, V. I., Maskevich, A. A., Kuzmitsky, V. A., Turoverov, K. K., and Kuznetsova, I. M. (2007) Computational study of thioflavin T torsional relaxation in the excited state. J. Phys. Chem. A. 111, 4829–4835 245. Freire, S., De Araujo, M. H., Al-Soufi, W., and Novo, M. (2014) Photophysical study of Thioflavin T as fluorescence marker of amyloid fibrils. Dye. Pigment. 110, 97–105 246. Nilsson, K. P. R., Lindgren, M., and Hammarström, P. (2018) Luminescent-conjugated oligothiophene probe applications for fluorescence imaging of pure amyloid fibrils and protein aggregates in tissues. in Methods in Molecular Biology, pp. 485–496, 1779, 485–496 247. Civitelli, L., Sandin, L., Nelson, E., Khattak, S. I., Brorsson, A. C., and Kågedal, K. (2016) The luminescent oligothiophene p-FTAA converts toxic Aβ1-42 species into nontoxic amyloid fibers with altered properties. J. Biol. Chem. 291, 9223–9243 248. Hellstrand, E., Boland, B., Walsh, D. M., and Linse, S. (2010) Amyloid β-protein aggregation produces highly reproducible kinetic data and occurs by a two-phase process. ACS Chem. Neurosci. 1, 13–18 249. Morris, A. M., Watzky, M. A., and Finke, R. G. (2009) Protein aggregation kinetics, mechanism, and curve- fitting: a review of the literature. Biochim. Biophys. Acta. 1794, 375–397 250. Meisl, G., Michaels, T. C. T., Arosio, P., Vendruscolo, M., Dobson, C. M., and Knowles, T. P. J. (2019) Dynamics and Control of Peptide Self-Assembly and Aggregation. Adv. Exp. Med. Biol. 1174, 1–33 251. Palmer, A. G. (1997) Probing molecular motion by NMR. Curr. Opin. Struct. Biol. 7, 732–737 252. Danielsson, J., Pierattelli, R., Banci, L., and Gräslund, A. (2007) High-resolution NMR studies of the zinc- binding site of the Alzheimer’s amyloid beta-peptide. FEBS J. 274, 46–59 253. Yamaguchi, T., Matsuzaki, K., and Hoshino, M. (2011) Transient formation of intermediate conformational states of amyloid-β peptide revealed by heteronuclear magnetic resonance spectroscopy. FEBS Lett. 585, 1097– 1102 254. Williamson, M. P. (2013) Using chemical shift perturbation to characterise ligand binding. Prog. Nucl. Magn. Reson. Spectrosc. 73, 1–16 255. Lindgren, J., Wahlström, A., Danielsson, J., Markova, N., Ekblad, C., Gräslund, A., Abrahmsén, L., Karlström, A. E., and Wärmländer, S. K. T. S. (2010) N-terminal engineering of amyloid-β-binding Affibody molecules yields improved chemical synthesis and higher binding affinity. Protein Sci. 19, 2319–2329 256. Palmer, A. G. (2014) Chemical exchange in biomacromolecules: Past, present, and future. J. Magn. Reson. 241, 3–17 257. Lian, L. Y., and Roberts, G. (2011) Protein NMR Spectroscopy: Practical Techniques and Applications, 10.1002/9781119972006 258. Clore, G. M. (2013) Seeing the invisible by paramagnetic and diamagnetic NMR. Biochem. Soc. Trans. 41, 1343–1354 259. Carr, H. Y., and Purcell, E. M. (1954) Effects of diffusion on free precession in nuclear magnetic resonance experiments. Phys. Rev. 94, 630–638 260. Meiboom, S., and Gill, D. (1958) Modified spin-echo method for measuring nuclear relaxation times. Rev Sci Instrum. 29, 688–691 261. Wang, C., Grey, M. J., and Palmer, A. G. (2001) CPMG sequences with enhanced sensitivity to chemical exchange. J. Biomol. NMR. 21, 361–366 262. Palmer, A. G., Kroenke, C. D., and Loria, J. P. (2001) Nuclear magnetic resonance methods for quantifying

71 microsecond-to-millisecond motions in biological macromolecules. Methods Enzymol. 339, 204–238 263. Korzhnev, D. M., and Kay, L. E. (2008) Probing invisible, low-populated states of protein molecules by relaxation dispersion NMR spectroscopy: An application to protein folding. Acc. Chem. Res. 41, 442–451 264. Abelein, A., Lang, L., Lendel, C., Gräslund, A., and Danielsson, J. (2013) Corrigendum to “Transient small molecule interactions kinetically modulate amyloid β peptide self-assembly” [FEBS Lett. 586 (2012) 3991- 3995]. FEBS Lett. 587, 1452–1452 265. Abelein, A., Gräslund, A., and Danielsson, J. (2015) Zinc as chaperone-mimicking agent for retardation of amyloid β peptide fibril formation. Proc. Natl. Acad. Sci. 112, 5407–5412 266. Schultz, S. G. (1961) Determination of the Effective Hydrodynamic Radii of Small Molecules by Viscometry. J. Gen. Physiol. 44, 1189–1199 267. Achuthan, S., Chung, B. J., Ghosh, P., Rangachari, V., and Vaidya, A. (2011) A modified Stokes-Einstein equation for Aβ aggregation. BMC Bioinformatics. 12, S13 268. Ruggeri, F. S., Habchi, J., Cerreta, A., and Dietler, G. (2016) AFM-Based Single Molecule Techniques: Unraveling the Amyloid Pathogenic Species. Curr. Pharm. Des. 22, 3950–3970 269. L. Lyubchenko, Y. (2015) Amyloid misfolding, aggregation, and the early onset of protein deposition diseases: insights from AFM experiments and computational analyses. AIMS Mol. Sci. 2, 190–210 270. Tong, Z., Mikheikin, A., Krasnoslobodtsev, A., Lv, Z., and Lyubchenko, Y. L. (2013) Novel polymer linkers for single molecule AFM force spectroscopy. Methods. 60, 161–168 271. Lv, Z., Roychaudhuri, R., Condron, M. M., Teplow, D. B., and Lyubchenko, Y. L. (2013) Mechanism of amyloid β-protein dimerization determined using single-molecule AFM force spectroscopy. Sci. Rep. 3, 2880 272. Lv, Z., Banerjee, S., Zagorski, K., and Lyubchenko, Y. L. (2018) Supported lipid bilayers for atomic force microscopy studies. in Methods in Molecular Biology, pp. 129–143, 1814, 129–143 273. Canale, C., Oropesa-Nuñez, R., Diaspro, A., and Dante, S. (2018) Amyloid and membrane complexity: The toxic interplay revealed by AFM. Semin. Cell Dev. Biol. 73, 82–94 274. Watanabe-Nakayama, T., and Ono, K. (2018) High-speed atomic force microscopy of individual amyloidogenic protein assemblies. in Methods in Molecular Biology, pp. 201–212, 1814, 201–212 275. Glusker, J. (1991) Structural aspects of metal liganding to functional groups in proteins. Adv. Protein Chem. 42, 1–76 276. Dudev, T., and Lim, C. (2014) Competition among metal ions for protein binding sites: Determinants of metal ion selectivity in proteins. Chem. Rev. 114, 538–556 277. Huang, E. (1997) Metal ions and synaptic transmission: Think zinc. Proc Natl Acad Sci U S A. 94, 13386–7 278. Branch, T., Barahona, M., Dodson, C. A., and Ying, L. (2017) Kinetic Analysis Reveals the Identity of Aβ- Metal Complex Responsible for the Initial Aggregation of Aβ in the Synapse. ACS Chem. Neurosci. 8, 1970– 1979 279. Lynch, T., Cherny, R., and Bush, A. (2000) Oxidative processes in Alzheimer’s disease: the role of Aβ-metal interactions. Exp. Gerontol. 35, 445–451 280. Atrián-Blasco, E., Gonzalez, P., Santoro, A., Alies, B., Faller, P., and Hureau, C. (2018) Cu and Zn coordination to amyloid peptides: From fascinating chemistry to debated pathological relevance. Coord. Chem. Rev. 371, 38– 55 281. Nikseresht, S., Bush, A. I., and Ayton, S. (2019) Treating Alzheimer’s disease by targeting iron. Br. J. Pharmacol. 176, 3622–3635 282. Squitti, R. (2012) Metals in alzheimer’s disease: a systemic perspective. Front Biosci (Landmark Ed). 17, 451– 72 283. Hureau, C., and Faller, P. (2009) Aβ-mediated ROS production by Cu ions: Structural insights, mechanisms and relevance to Alzheimer’s disease. Biochimie. 91, 1212–1217 284. Sadiq, S., Ghazala, Z., Chowdhury, A., and Büsselberg, D. (2012) Metal toxicity at the synapse: Presynaptic, postsynaptic, and long-term effects. J. Toxicol. 2012, 132671 285. Barnham, K. J., and Bush, A. I. (2008) Metals in Alzheimer’s and Parkinson's Diseases. Curr Opin Chem Biol. 12, 222–228 286. Barnham, K., and Bush, A. (2014) Biological metals and metal-targeting compounds in major neurodegenerative diseases. Chem Soc Rev. 43, 6727–49

72 287. Huat, T. J., Camats-Perna, J., Newcombe, E. A., Valmas, N., Kitazawa, M., and Medeiros, R. (2019) Metal Toxicity Links to Alzheimer’s Disease and Neuroinflammation. J. Mol. Biol. 431, 1843–1868 288. Lane, D. J. R., Ayton, S., and Bush, A. I. (2018) Iron and Alzheimer’s Disease: An Update on Emerging Mechanisms. J. Alzheimer’s Dis. 64, S379–S395 289. Mezzaroba, L., Alfieri, D. F., Colado Simão, A. N., and Vissoci Reiche, E. M. (2019) The role of zinc, copper, manganese and iron in neurodegenerative diseases. Neurotoxicology. 74, 230–241 290. Martins, A. C., Morcillo, P., Ijomone, O. M., Venkataramani, V., Harrison, F. E., Lee, E., Bowman, A. B., and Aschner, M. (2019) New insights on the role of manganese in alzheimer’s disease and parkinson’s disease. Int. J. Environ. Res. Public Health. 16, E3546 291. Squitti, R., Simonelli, I., Ventriglia, M., Siotto, M., Pasqualetti, P., Rembach, A., Doecke, J., and Bush, A. I. (2014) Meta-analysis of serum non-ceruloplasmin copper in Alzheimer’s disease. J. Alzheimer’s Dis. 38, 809– 822 292. Mocchegiani, E., Bertoni-Freddari, C., Marcellini, F., and Malavolta, M. (2005) Brain, aging and neurodegeneration: Role of zinc ion availability. Prog Neurobiol. 75, 367–390 293. Brewer, G. J. (2014) Alzheimer’s disease causation by copper toxicity and treatment with zinc. Front. Aging Neurosci. 6, 1–5 294. Religa, D., Strozyk, D., Cherny, R. A., Volitakis, I., Haroutunian, V., Winblad, B., Naslund, J., and Bush, A. I. (2006) Elevated cortical zinc in Alzheimer disease. Neurology. 67, 69–75 295. Maynard, C. J., Bush, A. I., Masters, C. L., Cappai, R., and Li, Q.-X. (2005) Metals and amyloid-beta in Alzheimer’s disease. Int. J. Exp. Pathol. 86, 147–159 296. Adlard, P. A., and Bush, A. I. (2006) Metals and Alzheimer’s disease. J. Alzheimer’s Dis. 10, 145–63 297. Zatta, P., Drago, D., Bolognin, S., and Sensi, S. L. (2009) Alzheimer’s disease, metal ions and metal homeostatic therapy. Trends Pharmacol. Sci. 30, 346–355 298. Popugaeva, E., Pchitskaya, E., and Bezprozvanny, I. (2018) Dysregulation of Intracellular Calcium Signaling in Alzheimer’s Disease. Antioxidants Redox Signal. 29, 1176–1188 299. Lovell, M. A., Robertson, J. D., Teesdale, W. J., Campbell, J. L., and Markesbery, W. R. (1998) Copper, iron and zinc in Alzheimer’s disease senile plaques. J. Neurol. Sci. 158, 47–52 300. Markesbery, W. R., Ehmann, W. D., Hossain, T. I. M., and Alauddin, M. (1984) Brain manganese concentrations in human aging and Alzheimer’s disease. Neurotoxicology. 5, 49–58 301. Li, Y., Jiao, Q., Xu, H., Du, X., Shi, L., Jia, F., and Jiang, H. (2017) Biometal dyshomeostasis and toxic metal accumulations in the development of alzheimer’s disease. Front. Mol. Neurosci. 10, 339 302. Solioz, M. (2020) Low copper-2 intake in Switzerland does not result in lower incidence of Alzheimer’s disease and contradicts the Copper-2 Hypothesis. Exp. Biol. Med. 245, 177–179 303. Cicero, C. E., Mostile, G., Vasta, R., Rapisarda, V., Signorelli, S. S., Ferrante, M., Zappia, M., and Nicoletti, A. (2017) Metals and neurodegenerative diseases. A systematic review. Environ. Res. 159, 82–94 304. Bihaqi, S. W. (2019) Early life exposure to lead (Pb) and changes in DNA methylation: Relevance to Alzheimer’s disease. Rev. Environ. Health. 34, 187–195 305. Bush, A. I. (2013) The Metal Theory of Alzheimer’s Disease. J. Alzheimers Dis. 33, S277–S281 306. Gaeta, A., and Hider, R. C. (2005) The crucial role of metal ions in neurodegeneration: the basis for a promising therapeutic strategy. Br. J. Pharmacol. 146, 1041–1059 307. Alies, B., Hureau, C., and Faller, P. (2013) The role of metal ions in amyloid formation: General principles from model peptides. Metallomics. 5, 183–192 308. Kenche, V. B., and Barnham, K. J. (2011) Alzheimer’s disease & metals: Therapeutic opportunities. Br. J. Pharmacol. 163, 211–219 309. Faller, P. (2009) Copper and zinc binding to amyloid-β: Coordination, dynamics, aggregation, reactivity and metal-ion transfer. ChemBioChem. 10, 2837–2845 310. Faller, P., Hureau, C., and La Penna, G. (2014) Metal ions and intrinsically disordered proteins and peptides: From Cu/Zn amyloid-β to general principles. Acc. Chem. Res. 47, 2252–2259 311. Moir, R. D., Atwood, C. S., Romano, D. M., Laurans, M. H., Huang, X., Bush, A. I., Smith, J. D., and Tanzi, R. E. (1999) Differential effects of apolipoprotein E isoforms on metal-induced aggregation of Aβ using physiological concentrations. Biochemistry. 38, 4595–4603

73 312. Lermyte, Everett, Brooks, Bellingeri, Billimoria, Sadler, O’Connor, Telling, and Collingwood (2019) Emerging Approaches to Investigate the Influence of Transition Metals in the Proteinopathies. Cells. 8, 1231 313. LeVine, H., Ding, Q., Walker, J. A., Voss, R. S., and Augelli-Szafran, C. E. (2009) Clioquinol and other hydroxyquinoline derivatives inhibit Aβ(1-42) oligomer assembly. Neurosci. Lett. 465, 99–103 314. Sampson, E. L., Jenagaratnam, L., and Mcshane, R. (2014) Metal protein attenuating compounds for the treatment of Alzheimer’s dementia. Cochrane Database Syst. Rev. 2014, CD005380 315. Budimir, A. (2011) Metal ions, Alzheimer’s disease and chelation therapy. Acta Pharm. 61, 1–14 316. Strodel, B., and Coskuner-Weber, O. (2019) Transition Metal Ion Interactions with Disordered Amyloid-β Peptides in the Pathogenesis of Alzheimer’s Disease: Insights from Computational Chemistry Studies. J. Chem. Inf. Model. 59, 1782–1805 317. Miller, L. M., Wang, Q., Telivala, T. P., Smith, R. J., Lanzirotti, A., and Miklossy, J. (2006) Synchrotron-based infrared and X-ray imaging shows focalized accumulation of Cu and Zn co-localized with β-amyloid deposits in Alzheimer’s disease. J. Struct. Biol. 155, 30–37 318. Everett, J., Collingwood, J. F., Tjendana-Tjhin, V., Brooks, J., Lermyte, F., Plascencia-Villa, G., Hands- Portman, I., Dobson, J., Perry, G., and Telling, N. D. (2018) Nanoscale synchrotron X-ray speciation of iron and calcium compounds in amyloid plaque cores from Alzheimer’s disease subjects. Nanoscale. 10, 11782–11796 319. Atwood, C. S., Huang, X., Khatri, A., Scarpa, R. C., Kim, Y. S., Moir, R. D., Tanzi, R. E., Roher, A. E., and Bush, A. I. (2000) Copper catalyzed oxidation of Alzheimer Abeta. Cell. Mol. Biol. 46, 777–783 320. Lyncha, T., Chernya, R., and Bush, A. (1999) Oxidative processes in Alzheimer’s disease: the role of Aβ-metal interactions. J. Environ. Monit. 35, 445–51 321. Wallin, C., Kulkarni, Y. S., Abelein, A., Jarvet, J., Liao, Q., Strodel, B., Olsson, L., Luo, J., Abrahams, J. P., Sholts, S. B., Roos, P. M., Kamerlin, S. C. L., Gräslund, A., and Wärmländer, S. K. T. S. (2016) Characterization of Mn(II) ion binding to the amyloid-β peptide in Alzheimer’s disease. J Trace Elem Med Biol. 38, 183–193 322. Bousejra-Elgarah, F., Bijani, C., Coppel, Y., Faller, P., and Hureau, C. (2011) Iron(II) binding to amyloid-β, the Alzheimer’s peptide. Inorg. Chem. 50, 9024–9030 323. Poulson, B. G., Szczepski, K., Lachowicz, J. I., Jaremko, L., Emwas, A. H., and Jaremko, M. (2019) Aggregation of biologically important peptides and proteins: Inhibition or acceleration depending on protein and metal ion concentrations. RSC Adv. 10, 215–227 324. Alies, B., Eury, H., Bijani, C., Rechignat, L., Faller, P., and Hureau, C. (2011) pH-dependent Cu(II) coordination to amyloid-β peptide: Impact of sequence alterations, including the H6R and D7N familial mutations. Inorg. Chem. 50, 11192–11201 325. Ghalebani, L., Wahlström, A., Danielsson, J., Wärmländer, S. K. T. S., and Gräslund, A. (2012) pH-dependence of the specific binding of Cu (II) and Zn (II) ions to the amyloid-β peptide. Biochem Biophys Res Commun. 421, 554–560 326. Eury, H., Bijani, C., Faller, P., and Hureau, C. (2011) Copper(II) coordination to amyloid β: Murine versus human peptide. Angew. Chemie - Int. Ed. 50, 901–905 327. Faller, P., and Hureau, C. (2009) Bioinorganic chemistry of copper and zinc ions coordinated to amyloid-beta peptide. Dalton Trans. 21, 1080–94 328. Atwood, C. S., Moir, R. D., Huang, X., Scarpa, R. C., Bacarra, N. M. E., Romano, D. M., Hartshorn, M. A., Tanzi, R. E., and Bush, A. I. (1998) Dramatic aggregation of alzheimer by Cu(II) is induced by conditions representing physiological acidosis. J. Biol. Chem. 273, 12817–12826 329. Dharmadana, D., Reynolds, N. P., Conn, C. E., and Valéry, Cé. (2017) Molecular interactions of amyloid nanofibrils with biological aggregation modifiers: Implications for cytotoxicity mechanisms and biomaterial design. Interface Focus. 7, 20160160 330. Ahrens, L. H. (1952) The use of ionization potentials Part 1. Ionic radii of the elements. Geochim. Cosmochim. Acta. 2, 155–169 331. Rayner-Canham, G. (2014) Descriptive Inorganic Chemistry 6e 332. Lemire, J. A., Harrison, J. J., and Turner, R. J. (2013) Antimicrobial activity of metals: Mechanisms, molecular targets and applications. Nat. Rev. Microbiol. 11, 371–384 333. De Gregorio, G., Biasotto, F., Hecel, A., Luczkowski, M., Kozlowski, H., and Valensin, D. (2019) Structural analysis of copper(I) interaction with amyloid β peptide. J. Inorg. Biochem. 195, 31–38

74 334. Lermyte, F., Everett, J., Lam, Y. P. Y., Wootton, C. A., Brooks, J., Barrow, M. P., Telling, N. D., Sadler, P. J., O’Connor, P. B., and Collingwood, J. F. (2019) Metal Ion Binding to the Amyloid β Monomer Studied by Native Top-Down FTICR Mass Spectrometry. J. Am. Soc. Mass Spectrom. 30, 2123–2134 335. Valensin, D., Migliorini, C., Valensin, G., Gaggelli, E., La Penna, G., Kozlowski, H., Gabbiani, C., and Messori, L. (2011) Exploring the reactions of β-amyloid (Aβ) peptide 1-28 with Al(III) and Fe(III) ions. Inorg. Chem. 50, 6865–6867 336. Pearson, R. G. (1963) Hard and Soft Acids and Bases. J. Am. Chem. Soc. 85, 3533–3539 337. Brännström, K., Ohman, A., Lindhagen-Persson, M., and Olofsson, A. (2013) Ca2+ enhances Aβ polymerization rate and fibrillar stability in a dynamic manner. Biochem. J. 450, 189–97 338. Weibull, M. G. M., Simonsen, S., Oksbjerg, C. R., Tiwari, M. K., and Hemmingsen, L. (2019) Effects of Cu(II) on the aggregation of amyloid-β. J. Biol. Inorg. Chem. 24, 1197–1215 339. Tõugu, V., Tiiman, A., and Palumaa, P. (2011) Interactions of Zn(II) and Cu(II) ions with Alzheimer’s amyloid- beta peptide. Metal ion binding, contribution to fibrillization and toxicity. Metallomics. 3, 250–261 340. Gaggelli, E., Janicka-Klos, A., Jankowska, E., Kozlowski, H., Migliorini, C., Molteni, E., Valensin, D., Valensin, G., and Wieczerzak, E. (2008) NMR studies of the Zn2+ interactions with rat and human β-amyloid (1-28) peptides in water-micelle environment. J. Phys. Chem. B. 112, 100–109 341. Innocenti, M., Salvietti, E., Guidotti, M., Casini, A., Bellandi, S., Foresti, M. L., Gabbiani, C., Pozzi, A., Zatta, P., and Messori, L. (2010) Trace copper(II) or zinc(II) ions drastically modify the aggregation behavior of Amyloid-β1-42: An AFM study. J. Alzheimer’s Dis. 19, 1323–1329 342. Hou, L., and Zagorski, M. G. (2006) NMR reveals anomalous copper(II) binding to the amyloid Aβ peptide of Alzheimer’s disease. J. Am. Chem. Soc. 128, 9260–9261 343. Rana, M., and Sharma, A. K. (2018) Cu and Zn interactions with Aβ peptides: consequence of coordination on aggregation and formation of neurotoxic soluble Aβ oligomers. Metallomics. 11, 64–84 344. Markesbery, W. R. (1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Radic. Biol. Med. 23, 134– 147 345. Al-Hilaly, Y. K., Williams, T. L., Stewart-Parker, M., Ford, L., Skaria, E., Cole, M., Bucher, W. G., Morris, K. L., Sada, A. A., Thorpe, J. R., and Serpell, L. C. (2014) A central role for dityrosine crosslinking of Amyloid-β in Alzheimer’s disease. Acta Neuropathol. Commun. 1, 83 346. Han, J. C., and Han, G. Y. (1994) A procedure for quantitative determination of tris(2-carboxyethyl)phosphine, an odorless reducing agent more stable and effective than dithiothreitol. Anal. Biochem. 220, 5–10 347. Han, J. (1996) Quantitation of Hydrogen Peroxide Using Tris(2-carboxyethyl)phosphine. Anal. Biochem. 234, 107–9 348. Huang, X., Atwood, C. S., Hartshorn, M. A., Multhaup, G., Goldstein, L. E., Scarpa, R. C., Cuajungco, M. P., Gray, D. N., Lim, J., Moir, R. D., Tanzi, R. E., and Bush, A. I. (1999) The Abeta peptide of Alzheimer’s disease directly produces hydrogen peroxide through metal ion reduction. Biochemistry. 38, 7609–7616 349. Durazzo, T. C., Mattsson, N., and Weiner, M. W. (2014) Smoking and increased Alzheimer’s disease risk: A review of potential mechanisms. Alzheimers. Dement. 10, S122–45 350. Guidotti, G., Brambilla, L., and Rossi, D. (2017) Cell-Penetrating Peptides: From Basic Research to Clinics. Trends Pharmacol. Sci. 38, 406–424 351. Raucher, D., and Ryu, J. S. (2015) Cell-penetrating peptides: Strategies for anticancer treatment. Trends Mol. Med. 21, 560–570 352. Ramsey, J. D., and Flynn, N. H. (2015) Cell-penetrating peptides transport therapeutics into cells. Pharmacol. Ther. 154, 78–86 353. Kokotidou, C., Jonnalagadda, S. V. R., Orr, A. A., Vrentzos, G., Kretsovali, A., Tamamis, P., and Mitraki, A. (2019) Designer amyloid cell-penetrating peptides for potential use as gene transfer vehicles. Biomolecules. 10, E7 354. Löfgren, K., Wahlström, A., Lundberg, P., Langel, Ö., Gräslund, A., and Bedecs, K. (2008) Antiprion properties of prion protein‐derived cell‐penetrating peptides. FASEB J. 22, 2177–2184 355. Yam, A. Y., Wang, X., Gao, C. M., Connolly, M. D., Zuckermann, R. N., Bleu, T., Hall, J., Fedynyshyn, J. P., Allauzen, S., Peretz, D., and Salisbury, C. M. (2011) A universal method for detection of amyloidogenic misfolded proteins. Biochemistry. 50, 4322–4329

75 356. Lau, A. L., Yam, A. Y., Michelitsch, M. M. D., Wang, X., Gao, C., Goodson, R. J., Shimizu, R., Timoteo, G., Hall, J., Medina-Selby, A., Coit, D., McCoin, C., Phelps, B., Wu, P., Hu, C., Chien, D., and Peretz, D. (2007) Characterization of prion protein (PrP)-derived peptides that discriminate full-length PrPSc from PrPC. Proc. Natl. Acad. Sci. U. S. A. 104, 11551–11556 357. Söderberg, K. L., Guterstam, P., Langel, Ü., and Gräslund, A. (2014) Targeting prion propagation using peptide constructs with signal sequence motifs. Arch. Biochem. Biophys. 564, 254–261 358. Schmitt-Ulms, G., Legname, G., Baldwin, M. A., Ball, H. L., Bradon, N., Bosque, P. J., Crossin, K. L., Edelman, G. M., DeArmond, S. J., Cohen, F. E., and Prusiner, S. B. (2001) Binding of neural cell adhesion molecules (N- CAMs) to the cellular prion protein. J. Mol. Biol. 314, 1209–1225 359. Santuccione, A., Sytnyk, V., Leshchyns’ka, I., and Schachner, M. (2005) Prion protein recruits its neuronal receptor NCAM to lipid rafts to activate p59fyn and to enhance neurite outgrowth. J. Cell Biol. 169, 341–354 360. Leshchyns’Ka, I., and Sytnyk, V. (2016) Synaptic Cell Adhesion Molecules in Alzheimer’s Disease. Neural Plast. 2016, 6427537 361. Karran, E., and De Strooper, B. (2016) The amyloid cascade hypothesis: are we poised for success or failure? J. Neurochem. 139, 237–252 362. Lippens, G., Landrieu, I., Smet, C., Huvent, I., Gandhi, N. S., Gigant, B., Despres, C., Qi, H., and Lopez, J. (2016) NMR meets Tau: Insights into its function and pathology. Biomolecules. 6, E28 363. Lippens, G., and Gigant, B. (2019) Elucidating Tau function and dysfunction in the era of cryo-EM. J. Biol. Chem. 294, 9316–9325 364. Panda, D., Samuel, J. C., Massie, M., Feinstein, S. C., and Wilson, L. (2003) Differential regulation of microtubule dynamics by three- and four-repeat tau: Implications for the onset of neurodegenerative disease. Proc. Natl. Acad. Sci. U. S. A. 100, 9548–9553 365. Mukrasch, M. D., Bibow, S., Korukottu, J., Jeganathan, S., Biernat, J., Griesinger, C., Mandelkow, E., and Zweckstetter, M. (2009) Structural polymorphism of 441-residue Tau at single residue resolution. PLoS Biol. 7, e34 366. Kolarova, M., García-Sierra, F., Bartos, A., Ricny, J., and Ripova, D. (2012) Structure and pathology of tau protein in Alzheimer disease. Int. J. Alzheimers. Dis. 2012, 731526 367. Huvent, I., Kamah, A., Cantrelle, F. X., Barois, N., Slomianny, C., Smet-Nocca, C., Landrieu, I., and Lippens, G. (2014) A functional fragment of Tau forms fibers without the need for an intermolecular cysteine bridge. Biochem. Biophys. Res. Commun. 445, 299–303 368. Bloom, G. S. (2014) Amyloid-β and tau: The trigger and bullet in Alzheimer disease pathogenesis. JAMA Neurol. 71, 505–508 369. Zempel, H., Luedtke, J., Kumar, Y., Biernat, J., Dawson, H., Mandelkow, E., and Mandelkow, E. M. (2013) Amyloid-β oligomers induce synaptic damage via Tau-dependent microtubule severing by TTLL6 and spastin. EMBO J. 32, 2920–2937 370. Frandemiche, M. L., De Seranno, S., Rush, T., Borel, E., Elie, A., Arnal, I., Lanté, F., and Buisson, A. (2014) Activity-dependent tau protein translocation to excitatory synapse is disrupted by exposure to amyloid-beta oligomers. J. Neurosci. 34, 6084–6097 371. Qu, X., Yuan, F. N., Corona, C., Pasini, S., Pero, M. E., Gundersen, G. G., Shelanski, M. L., and Bartolini, F. (2017) Stabilization of dynamic microtubules by mDia1 drives Tau-dependent Aβ1-42 synaptotoxicity. J. Cell Biol. 216, 3161–3178 372. Bilousova, T., Miller, C. A., Poon, W. W., Vinters, H. V., Corrada, M., Kawas, C., Hayden, E. Y., Teplow, D. B., Glabe, C., Albay, R., Cole, G. M., Teng, E., and Gylys, K. H. (2016) Synaptic Amyloid-β Oligomers Precede p-Tau and Differentiate High Pathology Control Cases. Am. J. Pathol. 186, 185–198 373. Pooler, A. M., Polydoro, M., Maury, E. A., Nicholls, S. B., Reddy, S. M., Wegmann, S., William, C., Saqran, L., Cagsal-Getkin, O., Pitstick, R., Beier, D. R., Carlson, G. A., Spires-Jones, T. L., and Hyman, B. T. (2015) Amyloid accelerates tau propagation and toxicity in a model of early Alzheimer’s disease. Acta Neuropathol. Commun. 3, 14 374. Tripathi, T., and Khan, H. (2020) Direct Interaction between the β-Amyloid Core and Tau Facilitates Cross- Seeding: A Novel Target for Therapeutic Intervention. Biochemistry. 59, 341–342 375. Saul, A., Sprenger, F., Bayer, T. A., and Wirths, O. (2013) Accelerated tau pathology with synaptic and neuronal

76 loss in a novel triple transgenic mouse model of Alzheimer’s disease. Neurobiol. Aging. 34, 2564–2573 376. Héraud, C., Goufak, D., Ando, K., Leroy, K., Suain, V., Yilmaz, Z., De Decker, R., Authelet, M., Laporte, V., Octave, J. N., and Brion, J. P. (2014) Increased misfolding and truncation of tau in APP/PS1/tau transgenic mice compared to mutant tau mice. Neurobiol. Dis. 62, 100–112 377. Chabrier, M. A., Cheng, D., Castello, N. A., Green, K. N., and LaFerla, F. M. (2014) Synergistic effects of amyloid-beta and wild-type human tau on dendritic spine loss in a floxed double transgenic model of Alzheimer’s disease. Neurobiol. Dis. 64, 107–117 378. Umeda, T., Maekawa, S., Kimura, T., Takashima, A., Tomiyama, T., and Mori, H. (2014) Neurofibrillary tangle formation by introducing wild-type human tau into APP transgenic mice. Acta Neuropathol. 127, 685–698 379. Guo, Q., Li, H., Cole, A. L., Hur, J. Y., Li, Y., and Zheng, H. (2013) Modeling alzheimer’s disease in mouse without mutant protein overexpression: Cooperative and independent effects of Aβ and tau. PLoS One. 8, e80706 380. Yetman, M. J., Fowler, S. W., and Jankowsky, J. L. (2016) Humanized tau mice with regionalized amyloid exhibit behavioral deficits but no pathological interaction. PLoS One. 11, e0153724 381. Guo, J. P., Arai, T., Miklossy, J., and McGeer, P. L. (2006) Aβ and tau form soluble complexes that may promote self aggregation of both into the insoluble forms in Alzheimer’s diseases. Proc. Natl. Acad. Sci. U. S. A. 103, 1953–1958 382. Penke, B., Szücs, M., and Bogár, F. (2020) Oligomerization and Conformational Change Turn Monomeric β- Amyloid and Tau Proteins Toxic: Their Role in Alzheimer’s Pathogenesis. Molecules. 25, E1659 383. Kellogg, E. H., Hejab, N. M. A., Poepsel, S., Downing, K. H., DiMaio, F., and Nogales, E. (2018) Near-atomic model of microtubule-tau interactions. Science. 360, 1242–1246 384. Mohamed, N. V., Herrou, T., Plouffe, V., Piperno, N., and Leclerc, N. (2013) Spreading of tau pathology in Alzheimer’s disease by cell-to-cell transmission. Eur. J. Neurosci. 37, 1939–1948 385. Frost, B., Jacks, R. L., and Diamond, M. I. (2009) Propagation of Tau misfolding from the outside to the inside of a cell. J. Biol. Chem. 284, 12845–12852 386. Avila, J. (2010) Intracellular and extracellular tau. Front. Neurosci. 4, 49 387. Reynolds, M. R., Berry, R. W., and Binder, L. I. (2005) Site-specific nitration differentially influences τ assembly in vitro. Biochemistry. 44, 13997–14009 388. Gamblin, T. C., Berry, R. W., and Binder, L. I. (2003) Modeling Tau Polymerization in Vitro: A Review and Synthesis. Biochemistry. 42, 15009–15017