Amyloid fibrillar aggregates of Aβ and tau are involved in Alzheimer disease. Short regions in these peptides trigger this aggregation. It is important to understand the basis of such short regions aggregation and amyloidosis for therapeutic intervention of Alzheimer disease. Set up datasets and use bioinformatics tool to data mining is an one way to explore diseases. Comparison of features of amyloidogenic segments with those of non-amyloidogenic segments could reveal factors that encode messages underlying the formation of amyloids aggregates. Such information combine with the mechanism of Alzheimer disease could be useful when designing a therapeutic regimen for prevent and treatment of Alzheimer diseases involving Aβ and tau aggregation, and provide basis for rational design of drugs for this neurodegenerative diseases. To investigate the properties of peptides and their effects on the amyloid fibril formation, we construct a comprehensive an amyloidogenic protein database, which include the information of the amyloidogenic segments and non-amyloidogenic segments. Based on a statistical study of this database, we arrived at the following conclusions. There is no significant difference in amino acid composition in sequence charcter, and for Aβ, such the discordants-containing amyloidogenic segments are observed with a more hydrophobic environment, which suggests that the hydrophobic environment might be one of the properties leading to the amyloid formation. We also analysis other specific physico-chemical properties of amyloidogenic segments and compare them with non-amyloidogenic segments. First, amyloidogenic segments are characterized by lower values for average net charge, electrostatic potential, solvent accessible surface area and B-factor when compared to the non-amyloidogenic segments of the same proteins. Second, they are enriched in hydrophobic residues and have a tendency to form hydrogen bonds. Thus, amyloidogenic segments have distinct physico-chemical properties that are different from those of non-amyloidogenic segments. Third, and quite unexpectedly, our dynamic simulation studies support the hypothesis that amyloidogenic segments have lower average flexibility than non-amyloidogenic segments. Furthermore, the presence of amyloidogenic segments in disordered proteins does not contradict the observation that amyloidogenic segments are less flexible. From evolustion study, there is almost no difference in evolutionary conservation between amyloidogenic segments and non-amyloidogenic segments. Beyond expec tation, rapidly evolving amino acid contribute to amyloid formation. In addition, the rapidly evolving amino acid is not evolve randomly, instead, they follow the same trend that with the uniform of physical characteristic. Our work demonstrated one aspect of such trend in evolution.
修改评论