Development And Application Of Force Field In Intrinsically Disordered Proteins

Recently, with the development of information era and the computation power, molecular dynamics(MD) simulation has been providing increasing support to the traditional “wet” researches in biomolecule structure-function relationship, pathogenesis and development of diseases, drug target searching, and drug design. As the fundamental basis of MD simulation, molecular force field plays a key role in the accuracy of MD simulation. Its optimization is also aiming for the improvement of MD simulation’s accuracy and applicability.Intrinsically disordered proteins(IDPs) belong to a type of proteins which lack specific spartial structures. The high flexibility of IDPs enables them to represent different structural characteristics and function from ordered proteins. Thus, IDPs are important in cell signaling pathway and associated with many diseases. Therefore, it is neccessay and urgent to understand the structure-function relationship of IDPs. However, existing force fields are not suitable for IDPs. In this thesis, the new IDP-specific force field, ff99 IDPs, is released via adding CMAP energy corrections for different residues to the dihedral energy term in ff99 SBildn. Evaluations and tests of ff99 IDPs on Me V NTAIL, p53 TAD, HIV-1 Rev protein, aspartic proteinase inhibitor IA3, and α-Synuclein show better performance on IDPs than ff99 SBildn. More disordered structures could be sampled, more accurate thermodynamic data could be predicted. Furthermore, tests on bound-Me V NTAIL, bound-HIVRev, lysozyme, and ubiquitin show that ff99 IDPs could also be used to simulate ordered proteins. However, under ff99 IDPs, over-stability in electrostatic and hydrogen bonding interactions could be found in some intrinsically disordered regions(IDRs), suggesting chargecharge interactions should be taken into consideration in next step. Therefore, in this study, CMAP energy corrections are also optimized and added to the polarizable force field ff02.pol.r1, which also yield significant better performance. Improvement and optimization are now proceeded on the polarizability and charge parameters.As an important transcription factor, ERG protein belongs to ETS-domain proteins family and is also associated with many diseases. Allosteric autoinhibition could be found in the binding between ERG protein and its target DNA. Many disordered regions could be observed in the crystal structure of ERG protein. In this study, multi-system and multi-trajectory MD simulations are performed for ERG-DNA complexes under ff99 IDPs, followed with correlation network analyses. Although the structures of autoinhibited and uninhibited ERG proteins are very similar, significant difference in correlation networks could be found between the two complexes. The difference is further confirmed to be due to the binding between N-terminal inhibitory dormain and C-terminal inhibitory domain. Thus, a hypothesis of “binding induced allostery” was proposed to explain the autoinhibition. Furthermore, this study suggests the need to augment the putative protein structure-function relationship with “structure �' dynamics network �' function” as a useful and maybe even general strategy in studies of protein structure-function relationship.