Applications of Elastic Network Model and Finite Element Analysis to Dynamics of Macromolecules

This dissertation addresses the applications of Elastic Network Model (ENM) and Finite Element Analysis (FEA) to molecular simulations on molecules and conformational dynamics of proteins. The applications include vibrational frequencies of molecules, the Revised Backbone ENM (RB-ENM), and conformational dynamics of transport factors in the Nucleocytoplasmic Transport (NCT). NCT is one of essential functions in cells, and structural changes or defects of associated macromolecules have been linked to a number of diseases such as breast cancer. The fundamental investigation on these macromolecules is needed to better understand the mechanism of NCT and the related diseases. Firstly, we used ENM to produce stretching force constants of linear molecules by matching their experimental vibrational frequencies and mode shapes. Later we used these generated force constants to predict the virbational frequencies of ethynyl isocyanide and diacetylene. Secondly, we introduced the RE-ENM to study conformations of a set of 17 open-closed proteins. The results showed that RE-ENM performed better on 12 out of 17 proteins than using traditional ENM. It makes RE-ENM a better scheme to study conformational changes of a protein. Finally, we used ENM and FEA to study conformational dynamics of two karyopherin proteins (Cse1p and Xpot), which play significant roles during the NCT. The proteins/exportins undergo distinguishing conformational changes to transport cargo molecules from the nucleus to the cytoplasm. We revealed the possible transitional pathways, and the result showed that the predicated intermediates were relevant to the known conformational changes. Furthermore, the conformational dynamics of exportins were interpreted by calculating intrinsic normal modes that played key roles in the flexibility of karyopherins, which further affected the binding affinities. The most interesting finding was that it was the karyopherin\u27s versatile conformations composed of the same superhelices of HEAT repeats that produced different degrees of functional flexibilities. The vibrational modes of exportins calculated by FEA were similar to the modes obtained from ENM. The distributions of strains helped to identify the possible hinge regions and quantified the flexibility of exportins. In the end, we presented the evidence that these coarse-grained methods could help to understand the biological function behind the structures of two transport factors