Intrinsic and extrinsic determinants of the aggregation process of amyloid proteins
Natalia Szulc · theses.fr (ABES) · 2023
Amyloids are a group of peptides and proteins that fold into assemblies of insoluble fibrils of very regular and tightly packed β-cross structures, which resemble a steric zipper. They are associated with many civilization diseases, such as type 2 diabetes, and wide range of neurodegenerative diseases. Amyloids' aggregation process and characteristics are greatly influenced by experimental conditions, which vary across studies and lack comprehensive information. This hampers the ability of bioinformatics methods to accurately predict amyloid propensity and structural models of aggregation. This dissertation investigates how internal and external factors impact the aggregation of amyloidogenic peptides, including sequence length, mutations, solvent properties, peptide interactions, and proximity to the cell wall. The study focuses on pathological and functional amyloids, utilizing both experimental and computational findings. First, we establish a reference protocol to investigate aggregation properties of short peptides for identifying amyloidogenic amino acid sequences. It assesses the impact of poorly annotated training data on bioinformatics methods' accuracy and explores the applicability of the protocol to longer sequences. We experimentally examined the amyloid propensity of long amino acid sequences (up to 23 amino acids) from the functional amyloid CsgA protein, specifically its R1-R5 imperfect fragments. It compared the sensitivity of these peptides to internal and external factors, such as point mutations, solvent, ion types, and ion concentrations, with a focus on homologous repeats from Escherichia coli and Salmonella enterica. Computational and theoretical studies validated the influence of mutations in this investigation. We finally examined the aggregation of pathological amyloids Aβ42 and hIAPP (sequences up to 42 amino acids), focusing on the impact of lipid membrane presence and peptide interaction on their aggregation under physiological conditions through experiments and simulations. The research findings emphasize the importance of collecting coherent data from experiments and highlight the influence of experimental conditions and sequence variations on amyloid aggregation. Short hexapeptide sequences exhibit distinct aggregation propensities, with flexibility and symmetry-breaking transitions playing a crucial role. The choice of solvent affects aggregation outcomes, as demonstrated by the R2 fragment's classification dependence on the presence of deuterium oxide. Amino acid substitutions and phosphate buffer ions also influence amyloid formation and morphology. Molecular dynamics simulations reveal the stability of Aβ42 within a lipid membrane and its interaction with hIAPP, impacting β-sheet content and supported lipid bilayer integrity. These results underscore the challenges in amyloid investigations and the need for considering experimental conditions and sequence differences in bioinformatics predictions. Effective modulation of peptide aggregation can be achieved through proper experimental planning and consideration of relevant conditions.