Computational Modeling of Beta-Fibrils

Abstract

Beta-sheet protein fibrils have been associated with many important pathological conditions such as Alzheimer's disease and type II diabetes. These fibrils are also important in the emerging field of peptide drug delivery since fibrillation is one of the critical degradation pathways for this class of drug molecules. Moreover, these structures have bionanotechnological applications such as formation of bio-nano-tubular scaffolds. Rational design of anti-fibrillation drugs, stable formulation of peptide drug molecules and development of bionanotechnological fibril devices all depend on understanding the structure of these fibrils. Although structures for some of such fibrils are experimentally determined, there are many others for which no structure is available. Hence, a computational method to model the structure of beta-sheet rich fibrils will be very valuable. We have devised such a computational method by combining homology modelling and MD simulation. The process starts with homology modelling of the beta-sheet building blocks using an optimized alignment and modelling scheme. Using those building blocks higher structures of the fibril are constructed. Finally, MD simulations are used to refine the fibril model and validate its energetic stability. We demonstrate the validity of our method by reproducing the experimentally determined structure of a number of beta-sheet fibrils. We also present the structure of glucagon beta-fibril for which no experimental structure is available. Understanding the structure of glucagon fibril can help inhibiting its fibrillation which is a major obstacle facing glucagon drug delivery.