Protein binding is an important issue in modern structure biology. People believe that proteins binding depend on their static and rigid structure. We found the evidence that the binding could also be determined by the flexibilit and dynamics of the structure. It's a charlenge to study the flexibility and dynamic property of proteins due to the limitation of the computer. I am using the simplified Go model.
Go model is based on coarse-grained molecular dynamics on the residue level with the energy function biased towards the native binding structure (Go model). With the model, the underlying free-energy landscape of the protein binding can be further explored.
We are going to extend our MD simulation to studying flexible binding of Camodulin signaling. We will carry out detailed simulations to explore the role of flexibility of CaM binding with the target peptides, using two-well model and Stochastic Difference Equation method.
In order to carry out a more detailed study of binding dynamics other than the simple lattice model, we will either start with the off lattice studies with the residue or atomic representations. As aforementioned, the detailed MD runs are still difficult due to the limitation of the computational power; we thus need to simplify the interaction energy in order to follow the whole dynamics of binding. The funneled shape of the landscape gives a hint of how to approximate the interactions. In addition to the usual energy terms such as bond, angles, torsion angles, non-bonded interactions (electrostatic and van der Waals interactions), a biasing (the Go potential) term towards native structure will be added. The rational of choosing the biasing term in the energy towards the native structure is that the binding mechanism seems to be largely determined by the topology of the native structure [9, 39, 40]
Go model is a simplified realization of landscape theory. It emphasizes the role of topology and native structures in determining the dynamic mechanisms. The energy function is chosen according to the native binding complex structure. In other words, the interaction energies among native contacts are uniformly attractive (often scale to value 1) stabilizing the native complex structure and the interactions of the non-native contacts are simply chosen to be zero value of energy (no bias towards the native structure). The potential can be implemented on lattice or off lattice simulations. In protein folding studies, people have implemented the Go potential for lattice, and off lattice simulations at the residue and atomic level. In this proposal, we will apply the atomic level Go potentials to study the binding-folding coupling dynamics by performing off lattice simulations. The advantage of the off lattice over lattice simulations is that the details such as the local secondary structures can be explicitly studied. The advantage of the atomic over residue level Go Model simulations is that the details such as local side chain packing can be explicitly studied. In our investigation, we will adapt the atomic Go model from our folding studies to binding and perform simulations taking into account of the microscopic details.