Molecular mechanisms in amyloid fibril formation

Abstract

This thesis concerns investigations on molecular processes and interactions that lead to amyloid fibril formation. Pulmonary surfactant protein C (SP-C), and a synthetic analogue thereof, SP-C(Leu), have similar properties in terms of charge distribution, hydrophobicity and secondary structure, yet they differ in their aggregation propensities.

These peptides were studied by hydrogen/deuterium exchange (HDX) MALDI mass spectrometry (MS). SP-C, but not SP-C(Leu) resists HDX and the ion current of SP-C but not of SP-C(Leu) was lost because of formation of insoluble aggregates. This suggests that SP-C has a high barrier of refolding to alpha-helix, responsible for the difference in aggregation ability of these similar peptides. A recently isolated biosynthetic precursor of SP-C, SP-Ci, has a 12-residue N-terminal propeptide followed by the SP-C sequence. It was observed that the propeptide locks the metastable SP-C in a helical conformation. The SP-Ci does not unfold or aggregate in neutral solution during several weeks of incubation, as judged by HDX and MS.

However, in an acidic environment SP-Ci unfolds and forms amyloid fibrils like SP-C. These data suggest a stabilizing role for the N-term inal propeptide in SP-C biosynthesis. Amyloid P-peptide (Abeta) is suggested to be responsible for Alzheimer's disease. The aggregation profile of Abeta(1-40) studied by electrospray ionisation (ESI) -MS, shows first order kinetics of Abetaaggregation. During the experiment, acid-catalysed spontaneous cleavage of Abeta(1 -40), most prominently at 23(Asp)-24(Val) was observed. Moreover, Abeta(24-40) was found to be unexpectedly stable in solution.

The tetrapeptides KFFE and KVVE can form fibrils, which are practically identical to fibrils formed in association with amyloid diseases. Other peptides studied, especially KLLE, KAAE, KFFK and EFFE did not form fibrils. In order to form fibrils the tetrapeptides thus need P-sheet promoting amino acids, and charge attraction from oppositely charged side chains. Attaching two KFFE sequences with either AAAK (supposedly random) or YNGK (frequent in betaturns) produces 12-mer peptides KFFEAAAKKFFE and KFFEYNGKKFFE. Freshly dissolved KFFEAAAKKFFE is monomeric and shows a random secondary structure, but upon incubation it forms abundant amyloid fibrils. Conversely, peptide KFFEYNGKKFFE does not form fibrils, but folds into a stable beta-hairpin structure. The results suggest that the tendency of polypeptides to form amyloid fibrils can depend on the structural context of an arnyloidogenic subsequence.