First Amyloid Crystal Structures Were Just The Beginning

Scientists used to think it wasn’t possible to crystallize and therefore learn the structure of amyloid fibrils, the protein complexes such as amyloid-β and α-synuclein that are involved in Alzheimer’s and Parkinson’s diseases, respectively. But David S. Eisenberg of the University of California, Los Angeles, and coworkers proved such thinking was wrong. In 2005, they crystallized and solved the structure of a segment of amyloid fibrils—called the cross-β spine—that holds the complexes together (Nature 2005, DOI: 10.1038/nature03680).

“We learned that the adhesive segments of the proteins that form these fibers are just short segments,” Eisenberg says. The researchers didn’t need to get the whole fiber to crystallize as long as they had the adhesive portion. They could then work with synthetic peptides instead.

The crystals formed by the segments were too small for conventional X-ray crystallographic methods. But synchrotron methods had luckily advanced enough that Eisenberg and his colleagues could analyze the tiny crystals. The first crystal structures revealed that the fibril spine consists of two β-sheets with interdigitating amino-acid side chains that form a zipperlike structure.

And those structures were just the beginning. Eisenberg’s team later showed that the steric zipper is a common structural motif in amyloid fibrils (Nature 2007, DOI: 10.1038/nature05695). In the past 10 years, they have solved the structures of the adhesive sections of many other disease-related proteins. This year, they solved the structure of the adhesive segment of α-synuclein, which had eluded them because it forms even smaller crystals than other segments do (Nature 2015, DOI: 10.1038/nature15368).

For Eisenberg, the crystal structures serve only as a jumping-off point. His team uses the atomic-level detail to design inhibitors that block fibril formation. Such inhibitors are typically peptides that bind to the top of a fibril and stop its growth. The first inhibitors were against tau, an Alzheimer’s-related protein, and SEVI (semen-derived enhancer of viral infection), a peptide that accelerates HIV infection (Nature 2011, DOI: 10.1038/nature10154). None of the inhibitors is yet in clinical trials, but the biotech firm ADRx has licensed the inhibitor-design method.

The 2005 work “was a breakthrough for amyloid structure studies and for crystallographic methodology,” says Robert Tycko, an amyloid expert at the National Institutes of Health. “So far, cross-β crystal structures of the full-length peptides and proteins have not been described, possibly because the inherent tendency of β-sheets to twist interferes with crystallization.” Studies of full-length fibril-forming proteins have been carried out with solid-state nuclear magnetic resonance spectroscopy instead.

“Crystallographers are notoriously successful in eventually crystallizing almost anything,” Tycko says. “So one should not rule out the possibility that cross-β structures of full-length amyloid-forming polypeptides will appear in the near future.”