Elucidating the structure of protein aggregates by Raman spectroscopy

The structures and properties of amyloid fibrils are of considerable interest due to their associations with numerous neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease and transmissible spongiform encephalopaties (prion diseases). Understanding fibrillogenesis at a molecular level requires detailed structural characterization of amyloid fibrils. However amyloid fibrils are difficult objects to study due to their non-crystalline and insoluble nature. These properties make the application of classical tools of structural biology, such as X-Ray crystallography and solution Nuclear Magnetic Resonance spectroscopy, impractical for structural characterization of protein fibrils.;Deep UV resonance Raman (DUVRR) spectroscopy is a powerful tool for protein structural characterization at all stages of aggregation and it does not require labeling of the peptide backbone. This unique advantage of Raman spectroscopy enables structural characterization of amyloid fibrils as well as other types of protein aggregates.;This work is dedicated to structural characterization of wide range of peptide aggregates including amyloid fibrils, novel non-fibrillar prion protein aggregates and fibril-mimicking peptide crystals by Raman spectroscopy. In the first part, the structural differences between the amyloid fibrillar core in parallel and anti-parallel cross-beta arrangements is analyzed using hydrogen-deuterium exchange UV Raman spectroscopy. In the second part, a mathematical model is created in order to obtain quantitative information about secondary structure content of protein aggregates. The model is based on advanced statistical analysis algorithms such as Partial Least Squares and Least Squares Support Vector Machines and includes a database of resonance Raman spectra of globular proteins together with amyloid fibrillar cores. This model can be used to obtain secondary structure information from proteins using their Raman spectra with 197 nm excitation. The model was used to calculate secondary structure composition of novel non-fibrillar aggregates of prion protein which have displayed positive amyloid features despite their non-fibrillar morphology. In the third part, the structures of fibrils and crystals of islet amyloid polypeptide fragments are characterized by Raman spectroscopy.