Limited Proteolysis Reveals That Amyloids from the 3D Domain-Swapping Cystatin B Have a Non-Native β-Sheet Topology

3D domain-swapping proteins form multimers by unfolding and then sharing of secondary structure elements, often with native-like interactions. Runaway domain swapping is proposed as a mechanism for folded proteins to form amyloid fibres, with examples including serpins and cystatins. Cystatin C amyl...

Full description

Saved in:
Bibliographic Details
Published inJournal of molecular biology Vol. 427; no. 15; pp. 2418 - 2434
Main Authors Davis, Peter J., Holmes, David, Waltho, Jonathan P., Staniforth, Rosemary A.
Format Journal Article
LanguageEnglish
Published England Elsevier Ltd 31.07.2015
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:3D domain-swapping proteins form multimers by unfolding and then sharing of secondary structure elements, often with native-like interactions. Runaway domain swapping is proposed as a mechanism for folded proteins to form amyloid fibres, with examples including serpins and cystatins. Cystatin C amyloids cause a hereditary form of cerebral amyloid angiopathy whilst cystatin B aggregates are found in cases of Unverricht-Lundborg Syndrome, a progressive form of myoclonic epilepsy. Under conditions that favour fibrillisation, cystatins populate stable 3D domain-swapped dimers both in vitro and in vivo that represent intermediates on route to the formation of fibrils. Previous work on cystatin B amyloid fibrils revealed that the α-helical region of the protein becomes disordered and identified the conservation of a continuous 20-residue elongated β-strand (residues 39–58), the latter being a salient feature of the dimeric 3D domain-swapped structure. Here we apply limited proteolysis to cystatin B amyloid fibrils and show that not only the α-helical N-terminal of the protein (residues 1–35) but also the C-terminal of the protein (residues 80–98) can be removed without disturbing the underlying fibril structure. This observation is incompatible with previous models of cystatin amyloid fibrils where the β-sheet is assumed to retain its native antiparallel arrangement. We conclude that our data favour a more generic, at least partially parallel, arrangement for cystatin β-sheet structure in mature amyloids and propose a model that remains consistent with available data for amyloids from either cystatin B or cystatin C. [Display omitted] •The cystatins are thought to form amyloid fibrils via a 3D domain-swapping mechanism.•Both the N-terminal 35 and the C-terminal 18 residues of cystatin B can be cleaved from fibrils without disturbing their structural integrity.•Cystatin B fibrils have a non-native topology incompatible with 3D domain swapping.•Cystatins are a good model for understanding amyloids in general.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0022-2836
1089-8638
DOI:10.1016/j.jmb.2015.05.014