Hydrogen bond models for the simulation of protein folding and aggregation

Abstract

Hydrogen bonds are a common feature in protein folding and aggregation. Due to their chemical peculiarities in terms of strength and directionality, a particular attention must be paid to the definition of the hydrogen bond potential itself. This global target has been tackled through a computational approach based on a minimalist description of the protein and the proper design of algorithms, mainly using Monte Carlo and Kinetic Monte Carlo methods. We have designed a hydrogen bond potential, see J. Chem. Phys. 132, 235102 (2010). We have been performed a complete study of sequenceless peptide systems under different conditions, such as temperature and concentration. To carry out full protein studies, we need additional potentials to describe tertiary interactions. We have discussed two different points of view. The first one is the combination of the hydrogen bond potential with a structure-based one. We have evaluated the implications of a proper definition of hydrogen bonds in the thermodynamic and dynamic aspects of protein folding. See Biophys. J. 101, 1474-1482 (2011). We have undertaken a second strategy, combining a generic hydrophobic model with the hydrogen bond one. The two main factors of folding and aggregation are merged, then, to create a simple but complete potential. Thanks to it, we have re-studied peptide aggregation including sequence. Besides, we have simulated complete proteins with different folded shapes. We have analyzed the competition between folding and aggregation, and how sequence and hydrogen bonds can influence the interplay between them. See J. Chem. Phys. 136, 215103 (2012).