Despite the first identification of proteinaceous amyloid fibrils as the defining feature of clinical disease over 50 years ago,1 the full structure of amyloid fibrils remains unresolved. Although more than 25 protein and peptide sequences are involved in human amyloid diseases2 or in the formation of functional amyloid fibrils in prokaryotes and eukaryotes,3 it is now widely accepted that most polypeptide sequences can assemble into amyloid fibrils under suitable experimental conditions.4,5 The resulting fibrils appear to be built of a common core structure in which two or more long ribbons of β-sheet (protofilaments) assemble into flat or twisted fibrils.6 Accordingly, amyloid fibrils can be identified by generic properties such as the ability to bind dyes like Congo red and thioflavin T, with the former resulting in the characteristic green birefringence of amyloid fibres.7 In addition, all amyloid fibrils give rise to characteristic low-frequency absorption bands indicative of highly organised β-sheets visualised by Fourier transform infrared (FTIR)8 and a cross-β fibre diffraction pattern indicating that the constituent β-strands are oriented perpendicular to the fibril long axis.9 Consistent with a generic fold for amyloid, all such fibrils are recognised by the antibody WO1 and other ligands commonly found associated with amyloid fibrils in disease,10,11 irrespective of the length or sequence of the constituent polypeptide chains.