Bioinformatics aggregation predictors in the study of protein conformational diseases of the human nervous system

Abstract

Conformational protein diseases of the human central nervous system represent a subject that has crucial theoretical and medical implications. They include several important neurodegenerative diseases, such as Alzheimer's, Parkinson's, Huntington's and Creutzfeldt-Jacob's diseases, amyotrophic lateral sclerosis, and the tauopathies. They occur when soluble proteins undergo conformational rearrangements becoming capable of aggregate into β-sheets conformations leading to the production of insoluble complexes known as amyloid deposits, that accumulate and lead to neurons and glial cells death. Theoretical and experimental evidence indicates that a key role in the conformational changes leading to amyloid formation is played by short sequence stretches within a given protein. Thus, the identification of protein regions potentially involved in aggregate formation and the characterization of their properties are relevant questions in the study of conformational proteins diseases. To address these questions, bioinformatics methods might provide an important contribution, suggesting possible mechanisms of protein aggregation, and focusing and orienting the experimental work. Thus, in the first part of the present review bioinformatics methods specifically attempting to predict aggregation-prone regions in proteins will be briefly described. Furthermore, the results provided by the combined use of some of them to analyze a set of particularly important proteins involved in human degenerative diseases will be discussed.