Relazione su invito

An unexpected journey in protein aggregation: From polymer physics to bioinformatic predictors and back again.

Trovato A.
  Giovedì 14/09   09:00 - 11:30   Aula A209   V - Biofisica e fisica medica
The interactions between different parts of a protein chain or between different protein chains are at the basis of their functional properties. However, the very same physico-chemical properties may also trigger the often detrimental association of proteins onto unsoluble aggregates. In particular, the formation of fibrillar amyloid aggregates upon protein misfolding is related to several devastating degenerative diseases in humans. On the other hand, the amyloid structure is also used by several organisms for their functional advantage. The basic amyloid motif of a cross-beta structure is common to many different sequences. At the same time different protein sequences exhibit varying propensities to aggregate into amyloids. In particular, how amyloidogenicity is enhanced by pathogenic mutations, the presence of aggregation hot spots stabilizing pathological interactions, the establishing of cross-amyloid interactions between co-aggregating proteins, all rely at the molecular level on the stability of the amyloid cross-beta structure. In this contribution I will first sketch how the "universal" features of the amyloid phase can be rationalized in a simple polymer physics framework. I will then discuss how a very fast and accurate predictor of aggregation propensity can be built by evaluating the pairing energy of two equal-length sequence stretches, when they are considered in a putative cross-beta arrangement. The method can be easily extended to evaluate the cross-beta pairings by two different sequences, allowing to predict on a genomic scale cross-amyloid interactions, and the co-aggregation of different proteins into heteromeric oligomer structures. I will end by briefly discussing the very recent evidence that liquid-liquid phase separation occurs for associating proteins $in vivo$ and a first attempt to capture the sequence-dependent properties of phase-separating proteins by means of simple mean-field lattice models.