Amyloid Peptide's Aggregation Mechanisms Studied By Molecular Dynamics Simulations

Proteins are the most abundant macromolecules in living cells. They form about 50% of the dry weight of a living cell. The structure of a protein determines its func-tion. Usually, a peptide or protein can fold correctly after the syntheses using RNA as a template. But sometimes they will misfold and even aggregate. A broad range of human diseases arise from the failure for a specific peptide or protein to fold into its native functional state. To explore the folding and aggregating mechanisms is im-portant for drug designing. With the development of computer techniques, molecular modeling has become a powerful tool to study the subjects related to macromolecules. From our study by using molecular dynamics methods, this thesis reports some mech-anisms existing in protein aggregation. These mechanisms include (1) the influnce of protein sequence on the agrregation; (2) the influence of single position mutate on the aggregation rate; (3) the mechanisms of nuleation and elongation, and (4) the factors that stabilize the structrue of fibrils. We use coarse grained OPEP model in order to study the former three mechanisms, while use all atom simulations to unveil the last. In detail, there are seven chapters:Chapter one gives a brief introduction to proteins, the interactions that stabilize proteins, some characteristics of protein folding and aggregation and some neurode-generative diseases.In the second chapter, we briefly introduce molecular dynamics methods (MD), force field and some basic analysis methods.In the third chapter, we report our results on two peptides, namely?2m83-89 and A?16-22, using OPEP model and constant temperature MD simulations. The two pep-tides contain the same number of amino acids, but have different hydrophobic prop-erties. Both of them contain 8 chains. The initial structure is random coil. Based on the free energy landscape analysis and association-dissociation analysis, we conclude that the hydrophobic property impact both the thermodynamics and dynamics of pro-tein aggregation. The higher hydrophobic characteristic, like A?16-22 with respect to?2m83-89, tends to decreases the size of oligomers by reducing the probability of bi-layer?-sheet; Also, the higher hydrophobic characteristic decreases the dynamics by reducing the association-dissociation rate, and by making it more difficult to transform between different conformations.In the forth chapter, we explore the impact of G33A and G33I mutant on the con-formational change of A?29-42 monomer and dimer using OPEP model and REMD simulations. From the simulations on monomers, We find that the G33A and G33I mutants decrease the probability of?-hairpin. From the simulations on dimers, we find that the G33A and G33I mutants decrease the intermolecular and long-range in-tramolecular interactions. Our results reveal that the mutants have a bigger probability to form amyloid-competent conformations, providing a plausible explanation for the faster aggregation of the two A?1-42 G33A and G33I mutants.In the fifth chapter, we explore the nucleation, elongation mechanisms related to protein aggregation by using OPEP model combined with both traditional MD and REMD. The system we use is NNQQ peptide which is the smallest amyloid-related peptide. Simulations were performed using 20 chains started from random arrange-ment. After analyzing the heat capacity with respect to temperature, the probability distribution of different oligomer size and the probability distribution of different?-sheet sizes, we find that there is a conformational transition around 333 K. This tran-sition is related to the breaking of Van der Walls interactions. The critical size of the nucleus might be 11-mer.Finally, by using all atom simulations we explore the impact of different side chains packing on the stability of the final product of aggregation (fibril). The system is NNQNTF peptide. It has two polymorphisms by micro-crystal techniques. We find that the factors that stabilize these two polymorphisms are different: for the back-to-back conformation, the PHE-packing is more important to keep the conformation in a good arrangement, while the face-to-face conformation prefers to form inter-sheet hy-drogen bonds. These two kinds of stabilizing factors are under the influence of water molecules. Also, we find that the state with water molecules in between the two layers maybe a intermediate state of the final fibril.The last chapter summarizes the above contents briefly and gives a perspective viewpoint on the field of protein aggregation.