Biophysical studies of b2-microglobulin amyloid formation

David Smith, Victoria McParland, Susan Jones, Neil Kad, Sheena Radford

Introduction

Amyloid disease involves the association of protein or peptide monomers into long, non-covalently associated protein fibrils. These fibrils can be formed from a wide range of normally soluble proteins and peptide fragments in vitro, and about 20 different human amyloid diseases are currently known. Amyloid fibrils share a similar cross-b structure, involving repeating b-strands, which lie perpendicular to the fibril long axis.

Work in our laboratory focuses on the human disease, haemodialysis-related amyloidosis which involves the aggregation of wild-type human b2-microglobulin (b2m) into amyloid fibrils. b2m forms the light chain of the cell surface class I MHC complex. The protein is shed continuously from the cell surface and is carried in the serum to the kidneys where it is catabolised and excreted. Upon kidney dysfunction b2m is no longer cleared from the serum in this manner and, as a consequence, serum levels increase by up to 20-fold, resulting in the formation and deposition of b2m amyloid in the joints. Like other amyloid diseases, the mechanism of b2m amyloidosis is currently unknown. Our laboratory is currently using structural and kinetic methods to determine the conformation of amyloidogenic intermediates, the mechanism of fibril assembly, and the structure of the amyloid fibril itself. This work is essential in order to provide a molecular understanding of amyloidosis and as a prelude to deriving therapies against this, and other, amyloid diseases.

Factors affecting amyloid formation

b2m fibrils have been visualised in macrophage lysosomes suggesting that this compartment could be involved in amyloidosis in vivo. In accord with this hypothesis, we have shown that a reduced pH environment is crucial for the formation of b2m amyloid fibrils in vitro. At neutral pH, b2m is native, stable, and fibril formation is not seen even after incubation for months. By contrast, fibril formation is rapid below pH 5.0. Use of the amyloid-specific dye, thioflavin-T, has allowed us to determine the initial rate of fibril formation over a range of pH values. These results showed that fibril formation is rapid at pH 3.6, whereas at lower pH fewer nucleation sites develop and fibril formation is slowed. Our studies in vitro have also shown that ionic strength plays a key role in fibril formation. At pH 3.6 the rate of fibrillogensis is increased markedly as the ionic strength is increased. Using negative stain electron microscopy, we have also shown that the conditions under which fibrils are grown affect their morphology. At pH 3.6 in high salt, the fibrils formed are short and curvilinear, whilst at lower pH the fibrils extend to much greater lengths.

The conformation of the monomeric amyloid precursor of b2m

At low ionic strengths the rate of b2m amyloidosis is very slow. Making use of this observation, we have found conditions under which the conformational properties of the monomeric amyloid precursor can be determined. Using circular dichroism, ANS binding, hydrogen exchange and 1H NMR, we have shown that the amyloid precursor is partially folded in that it retains substantial b-sheet structure, lacks fixed tertiary interactions and is weakly protected from hydrogen exchange. Detailed titration of the protein using far UV CD has shown that one or more groups with an apparent pKa of 4.7 are involved in the formation of the amyloid precursor state (figure 1). Current work is aimed at extending these studies by the use of multidimensional heteronuclear NMR methods and site-directed mutagenesis to derive residue-specific information about the conformational properties of the amyloid precursor state. In addition, FTIR is being used to determine information about the conformational changes occurring during fibril assembly.

Examining fibrillar intermediates

In parallel with the studies described above, we are using atomic force microscopy and mass spectrometry to examine the structure of oligomeric species populated during fibril assembly. This will allow us to discover whether b2m fibrillogenesis follows a nucleation dependent pathway or assembles through a ladder of intermediates. Studies of other amyloid diseases have suggested that small oligomeric amyloid precursors could be the key pathogenic agents in disease and hence it is critical to examine the occurrence and structure of these species in the search for new therapeutic agents.

Collaborators

Neil Thomson, Department of Physics and Astronomy and Astbury Centre for Structural Biology, University of Leeds.

Arnout Kalverda and Steve Homans, Astbury Centre for Structural Biology, University of Leeds.
Alison Ashcroft, Astbury Centre for Structural Biology, University of Leeds.
Sandy Davison, St. James’ Hospital, Leeds.
Margie Sunde, Department of Biochemistry, University of Cambridge.
Mick Hunter and Ant Brown, British Biotech Pharmaceuticals Ltd., Oxford.

References

Aggeli, A., Bell, M., Boden. N., Harding, R., McLeish, T.C.B., Nykova, I., Radford, S.E. and Semenov, A. (2000) Exploiting protein-like self-assembly to engineer nanostructured materials. (2000), The Biochemist 22, 10-15.

McParland V.J., Kad, N.M., Kalverda, A.P., Brown, A., Kirwin-Jones, P., Hunter, M.G., Sunde, M., Radford, S.E. (2000) Partially unfolded states of b2-microglobulin and amyloid formation in vitro. Biochemistry 39, 8735-8746.

Kad, N.M. and Radford, S.E. (2000) Partial unfolding as a precursor to amyloidosis: A discussion of the occurrence, role and implications. Frontiers in Molecular Biology. Ed. P. Lund, Oxford University Press. In press.

Funding

We thank British Biotech Pharmaceuticals Ltd, BBSRC, EPSRC and the Wellcome Trust for funding.