This work investigates the influence of mutations at selected positions on the structure formation of the Alzheimer’s disease peptide amyloid β. Amyloid β is a member of the class of intrinsically disordered proteins that can aggregate into fibrils, which are characterized by a highly stable secondary structure, called cross-β structure. A central contact during fibrillation is the hydrophobic F19-L34 contact, which is located within the core of the cross-β structure. Modifications of this contact are known to influence the local molecular structure whereas the fibril morphology and the cross-β structure remain stable. In contrast, toxicity of amyloid β was completely lost for all previously investigated mutants of F19 and L34. This work characterizes the properties of this contact and answers the question what the minimally tolerated modifications are. To characterize the structure, structure formation process and biological activity of the Aβ variants a set of experiments was carried out. The local structure and dynamics were investigated using NMR experiments focusing on 13C-chemical shift changes and 1H-13C dipolar couplings, respectively. The fibril morphology and cross- β structure was verified by electron microscopy, circular dichroism spectroscopy and X-ray diffraction. Toxicity and biological activity was investigated using complementary cell culture experiments. The work was divided in three parts. First, L34 was substituted with three highly similar amino acids: the isomer isoleucine, valine that is one methylene group shorter but also a branched chain amino acid and the stereoisomer D-leucine. The L34 position proved to be important for the initiation of the structure formation, oligomer stability, fibril growth and the biological activity of amyloid β. These characteristics and properties were highly sensitive also to minor modifications but the different mutants showed no specific but qualitatively similar effects. The second part complemented previous mutation studies of the F19 position. Four new mutants were designed testing mild modification of the F19-L34 contact: phenylglycine and the homophenylalanine (S)-2-amino-4-phenyl-butyric acid change the length of the side chain, cyclohexyl-alanine eliminates the π-aromaticity of the ring system and increases the 3D steric demand, and (1-naphtyl)-alanine increases the 2D steric demand while maintaining the aromaticity. Mutations at the F19 position caused qualitatively similar effects as L34 modifications but proved to have quantitatively greater impact. Furthermore, they showed some specificity as steric constraints caused larger changes than modifications of the ring system. The third part investigates the influence of β-methylamino-L-alanine (BMAA) substitutions at positions F19, S8, and S26. The serine to BMAA substitutions were included because of their potential medical relevance. A F19BMAA substitution caused similar effects like other modifications at this position. Replacement of serine lead to a structural reorientation of the Aβ N-terminus and turn region. Furthermore, the pathways of the cell response changed from mitochondrial activity and plasma membrane integrity to apoptosis and neuronal stress reaction. Summarizing, it could be shown that, although the formation and structure of amyloid β fibrils is robust against different modifications the fibrillation kinetics, local structure and especially biological activity is highly sensitive and to some extend specific to even minor modifications.

It is suggested that Alzheimer’s disease (AD) is caused by an imbalance between production, degradation and clearance of the amyloid-β (Aβ) protein. This imbalance leads to aggregation of Aβ and tau proteins and neurodegeneration in the brain. Today there is increasing evidence that the balance between the protease cathepsin B and the protease inhibitor cystatin C affects the tendency for Aβ to aggregate. The primary aim of this thesis was to investigate Aβ, cystatin C and cathepsin B levels in blood and cerebro-spinal fluid (CSF) in relation to the risk of AD. Studies I & II were based on the re-examinations of participants, at ages 70 and 77, in the Uppsala Longitudinal Study of Adult Men (ULSAM), a community-based prospective study initiated in 1970 (participants then being 50 years of age). In ULSAM, low plasma Aβ1-40 (Study I) and low serum cystatin C levels (Study II) were associated with a higher risk of AD. Studies III & IV were based on a cross-sectional sample of people with AD, mild cognitive impairment and healthy controls, recruited at three Swedish Memory Disorder units: Uppsala University Hospital, Uppsala, Skåne University Hospital, Malmö, and Karolinska University Hospital, Huddinge, Stockholm. In Study III, CSF cystatin C levels were positively correlated with both Aβ1-42 and tau levels. In Study IV, individuals with AD had higher mean plasma cathepsin B levels than healthy controls. In conclusion, low plasma Aβ1-40 and low serum cystatin C levels may precede clinically manifest AD in elderly men, cystatin C levels are positively correlated with Aβ1-42 and tau levels in CSF, and mean plasma cathepsin B levels are higher in people with AD compared to healthy controls. In addition to Aβ1-42 and tau levels in CSF, Aβ1-40, cystatin C and cathepsin B levels in blood may reflect the risk of AD.