Aggregation propensity, amyloid peptide toxicity, and Alzheimer Disease

When polypeptides fail to achieve or maintain their proper conformation, they commonly aggregate into abnormal “amyloid fibril” structures. Amyloid fibrils define a diverse group of degenerative conditions, including amyotrophic lateral sclerosis, prion diseases, and Alzheimer and Parkinson diseases.

In Alzheimer disease, the amyloid fibrils are deposited extracellularly; however, in Parkinson and Huntington disease, similar amyloid fibrils accumulate in the cytoplasm and nucleus of the cell respectively. How amyloid formation promotes disease has generated considerable debate, though mounting evidence implicates the early protofibrillar aggregates as the toxic species.

In a new study in the open-access journal PLoS Biology, Leila Luheshi et al. worked with the fruit fly Drosophila to identify the intrinsic determinants of amyloid ß (Aß) pathogenicity in an animal model of Alzheimer disease. (Aß peptide is a primary component of amyloid plaques in the brains of patients with Alzheimer disease.) Determining how amyloid formation causes disease requires a better understanding of the molecular and biophysical conditions that promote protein aggregation. But such an understanding has proven technically challenging, in part because protein misfolding and aggregation in test tubes can’t replicate cellular pathways designed to mitigate the toxic effects of these events. Luheshi et al. circumvented this problem by integrating computational predictions of protein aggregation propensities with in vitro experiments to test the predictions and in vivo mutagenesis experiments to link predicted aggregation propensity with observed neurodegeneration in the flies.

Overall, the researchers found a clear correlation between a variant’s predicted tendency to aggregate and its influence on fly longevity. The same relationship was seen between predicted aggregation propensity and locomotion, though a few variants did not follow this pattern. An interesting case presented with a variant (131E/E22G), whose neuronal effects did not match its predicted aggregation propensity. The 131E/E22G peptide aggregated at rates similar to the Alzheimer variant in vitro as well as in the fly brains. But because the 131E/E22G peptide deposits were not accompanied by cavities in brain tissue—a telltale sign of neurodegeneration—the flies showed no neurological deficits.

This finding fits with reports that the density of Aß plaques in elderly patients with Alzheimer disease does not correlate with the severity of clinical symptoms. Instead, it is the soluble protofibrillar aggregates, not the mature amyloid plaques, that cause neurodegeneration. Recomputing the propensities of each Aß variant to form these protofibrillar species revealed not only an improved overall correlation with toxicity, but it also brought the previously anomalous 131E/E22G variant in line with the prediction algorithm.

Altogether, these results show that Aß’s toxic effects in a living organism can be predicted based on a computational analysis of its tendency to form protofibrillar aggregates. And even though cells have evolved multiple mechanisms to regulate folding, the researchers argue, it is the intrinsic tendency of the peptide’s sequence to aggregate that governs its pathological propensity. Though the researchers focused on the peptide most closely associated with Alzheimer disease, they believe their approach will work for many other diseases as well.