Yeast Approach in Amyloidogenic Proteins Study ^†

Self-reproducing protein aggregates of a fibrillar nature (amyloids) are associated with a number of animal and human diseases. Some amyloids (prions) possess infectious properties as they can be transmitted between tissues, organs, and even organisms. However, there are examples of prions and amyloids that have normal biological functions [1]. The number of proteins capable of forming functional and pathological amyloids is constantly expanding. For example, some bioinformatics algorithms predict that over a hundred human proteins may have the ability to form amyloids.

In our study, we applied the ArchCandy algorithm to identify human proteins with amyloidogenic properties. This algorithm is capable of identifying the potential formation of β-arches (structures specific to amyloid proteins). By applying an in silico approach, we selected nine proteins that demonstrated amyloidogenic potential and experimentally validated these predictions. To evaluate the amyloidogenic potential of proteins, we employed an original yeast test system that we developed [2]. This system enables us to assess amyloidogenicity of individual proteins as well as to conduct large-scale screenings of amyloidogenic proteins using a genetic analysis. Using this test system, we demonstrated that six out of the nine predicted amyloidogenic proteins were prone to amyloid formation. Among the proteins that demonstrated amyloidogenic properties, there were proteins involved in chromatin remodeling, transcription regulation, and oncogenesis. For several potentially amyloidogenic proteins identified in the yeast model, their amyloid properties were confirmed both in vitro and in a human cell culture.