Nanomechanical properties of amyloid fibrils and nanocrystals depend on their secondary and quaternary structure, and the geometry of intermolecular hydrogen bonds. Advanced imaging methods based on atomic force microscopy (AFM) have unravelled the morphological and mechanical heterogeneity of amyloids, however a full understanding has been hampered by the limited resolution of conventional spectroscopic methods. Here, it is shown that single molecule nanomechanical mapping and infrared nanospectroscopy (AFM-IR) in combination with atomistic modelling enable unravelling at the single aggregate scale of the morphological, nanomechanical, chemical, and structural transition from amyloid fibrils to amyloid microcrystals in the hexapeptides, ILQINS, IFQINS, and TFQINS. Different morphologies have different Young's moduli, within 2-6 GPa, with amyloid fibrils exhibiting lower Young's moduli compared to amyloid microcrystals. The origins of this stiffening are unravelled and related to the increased content of intermolecular beta-sheet and the increased lengthscale of cooperativity following the transition from twisted fibril to flat nanocrystal. Increased stiffness in Young's moduli is correlated with increased density of intermolecular hydrogen bonding and parallel beta-sheet structure, which energetically stabilize crystals over the other polymorphs. These results offer additional evidence for the position of amyloid crystals in the minimum of the protein folding and aggregation landscape.

Several atomic structures have now been found for micrometer-scale amyloid fibrils or elongated microcrystals using a range of methods, including NMR, electron microscopy, and X-ray crystallography, with parallel β-sheet appearing as the most common secondary structure. The etiology of amyloid disease, however, indicates nanometer-scale assemblies of only tens of peptides as significant agents of cytotoxicity and contagion. By combining solution X-ray with molecular dynamics, we show that antiparallel structure dominates at the first stages of aggregation for a specific set of peptides, being replaced by parallel at large length scales only. This divergence in structure between small and large amyloid aggregates should inform future design of molecular therapeutics against nucleation or intercellular transmission of amyloid. Calculations and an overview from the literature argue that antiparallel order should be the first appearance of structure in many or most amyloid aggregation processes, regardless of the endpoint. Exceptions to this finding should exist, depending inevitably on the sequence and on solution conditions.

By combining atomistic and higher-level modelling with solution X-ray diffraction we analyse self-assembly pathways for the IFQINS hexapeptide, a bio-relevant amyloid former derived from human lysozyme. We verify that (at least) two metastable polymorphic structures exist for this system which are substantially different at the atomistic scale, and compare the conditions under which they are kinetically accessible. We further examine the higher-level polymorphism for these systems at the nanometre to micrometre scales, which is manifested in kinetic differences and in shape differences between structures instead of or as well as differences in the small-scale contact topology. Any future design of structure based inhibitors of the IFQINS steric zipper, or of close homologues such as TFQINS which are likely to have similar structures, should take account of this polymorphic assembly.

Amyloidogenic model peptides are invaluable for investigating assembly mechanisms in disease related amyloids and in protein folding. During aggregation, such peptides can undergo bifurcation leading to fibrils or crystals, however the mechanisms of fibril-to-crystal conversion are unclear. We navigate herein the energy landscape of amyloidogenic peptides by studying a homologous series of hexapeptides found in animal, human and disease related proteins. We observe fibril-to-crystal conversion occurring within single aggregates via untwisting of twisted ribbon fibrils possessing saddle-like curvature and cross-sectional aspect ratios approaching unity. Changing sequence, pH or concentration shifts the growth towards larger aspect ratio species assembling into stable helical ribbons possessing mean-curvature. By comparing atomistic calculations of desolvation energies for association of peptides we parameterise a kinetic model, providing a physical explanation of fibril-to-crystal interconversion. These results shed light on the self-assembly of amyloidogenic peptides, suggesting amyloid crystals, not fibrils, represent the ground state of the protein folding energy landscape.