Multiscale Simulations of Intrinsically Disordered Proteins [thesis]

Xiaorong Liu
Intrinsically disordered proteins (IDPs) lack stable secondary and/or tertiary structures under physiological conditions. The have now been recognized to play important roles in numerous biological processes, particularly cellular signaling and regulation. Mutation of IDPs are frequently associated with human diseases, such as cancers and neuron degenerative diseases. Therefore, it is important to understand the structure, dynamics, and interactions of IDPs, so as to establish the mechanistic
more » ... h the mechanistic basis of how intrinsic disorder mediates versatile functions and how such mechanisms may fail in human diseases. However, the heterogeneous structural ensembles of IDPs are not amenable to high resolution characterization solely through experimental measurements, and molecular modelling and simulation are required to study IDP structures, dynamics, and interactions at the atomistic levels. Here, we first applied the state-of-the-art explicit solvent atomistic simulations to an anti-apoptotic protein Bcl-xL and demonstrated how inherent structural disorder may provide a physical basis of protein regulated unfolding in signaling transduction. We have also constructed a series of efficient coarse-grained models to directly simulate the interactions between IDPs and unveiled how the preexisting structural elements accelerate binding of ACTR to NCBD by promoting efficient folding upon encounter. These studies shed important light on how IDPs perform functions in the cellular regulatory network, but also reveal the necessity of new sampling techniques for more efficient simulations of IDPs. We have thus developed a novel sampling technique, called multiscale enhanced sampling (MSES). MSES couples the atomistic model with coarse-grained ones, to accelerate the sampling of atomistic conformational space. Bias from coupling to a coarse-grained model can be removed using Hamiltonian replica exchange. To achieve the best possible efficiency of MSES simulations, we have developed a new hybrid resolution protein model that could captur [...]
doi:10.7275/14020756 fatcat:ctjqdaoy2jenhorrgsotlekfwe