Molecular Dynamics Simulation On Coupled Binding-folding And Conformational Change Of Proteins

The structures and function of biomolecules have a closed relationship.Proteins function is more complicate due to the dynamics of structure.But these structure dynamics cannot be exactly measured through experiments.These structures are usually measured in static state in crystal form,or just some special local interactions in dynamic form.Most of these experimental results are ensemble averaged data in some observed quantity.Single-molecule methods can deal with the detail local interaction pairs and trace the motion with time evolution.But if the whole molecule dynamics should be traced,much more interaction pairs should be marked and traced.This will make it difficult to observe at the same time,and too many label molecules will influence the target big protein dynamics.Molecular dynamics simulations can provide the details of large molecules conformational change and interactions.Once the simulation trajectories are generated,all the integral or local observed quantities can be extracted from the trajectories.In this work,we’ll present three examples on long protein binding in different flexibilities,coupled binding-folding of nucleoprotein segment and target,and conformational change of T4 lysozyme in ligand binding.In the first section,an intrinsically disordered protein segment Chz.core coupled binding-folding mechanism with histone variant complex H2 A.z-H2 B.We provide the atomistic structure-based model for the simulations of the protein complexes.Previously,we used C-α/C-β level structure based model and give a detailed coupled binding-folding mechanism.In this work,due to the integral backbone and side-chain description in all-atom level,we tuned the model to fit with experimental data,and the side-chain behavior in the binding process.Since intrinsically disordered protein binding mechanism is not as clear as ordered ones,and the efficiency in recognizing and binding compared with ordered proteins also need much more detection.So we give a systematic comparison of disordered and ordered proteins in a series of flexibilities.This is performed by tuning the intra-chain strength of Chz.core to mimick proteins in different flexibilities.The binding free energy landscape profiles in different flexible systems show that proteins with higher flexibility take lower binding barrier.This means flexible proteins binding will take lower energy,and easer to bind with partners.Meanwhile,we tested the kinetics of the series of proteins in different flexibilities.The total binding process can be separate into the capture process and the final binding process.In the mean passage time of capture test,we found that proteins with higher flexibilities take shorter time to touch the target H2 A.z-H2 B.On the other hand,it cannot be neglected that most capture events cannot fall to the bound state due to the weak interactions between two proteins.Proteins with higher flexibilities also take much less capture events to get to the final binding state.That is to say,they have higher conversion efficiency of capturing to binding.In the final binding process,our model show flexible proteins bind faster than ordered ones.All these results support that intrinsically disordered proteins take higher capturing and binding efficiency.The second protein binding systems are from the Hendra virus and the Nipah virus nucleoproteins C terminal tail segment binding with target proteins.These two virus are both belong to dangerous genus,and their genes are capsidated by the nucleoproteins.Their functions must contain the process of binding with the X domain of phosphoproteins,so we provided an atomic level structure-based simulation for the coupled binding and folding process.This model was successfully provide the protein complexes form measles virus,the same genus virus with Hendra and Nipah virus,so we take it as a reference for our simulations.Since the nucleoprotein segments from both viruses are intrinsically disordered proteins,the unfolded state cannot be well simulated by the structure-based models since they take the native order structure as the force field center and be less accuracy when proteins structures stay far away from the force field center.So we used the developed model hybrid with structure-based non-local interactions and local empirical force-field based interactions for the flexible binding process.First,we simulated the nucleoproteins at the isolated state in solutions with empirical force-field based model.This will detect the underlying structures and folding-unfolding mechanisms of these intrinsically disordered nucleoprotein segments.We used replica exchange methods to accelerate the conformational sampling,and give the most probable structures among the conformational probability distribution.In the coupled binding-folding process,the hybrid model show that protein binding from Hendra virus takes lower binding barrier and the binding free energy profile is alike with that from Nipah virus.But both of them take lower binding barrier than the measle virus.In the early binding stage,the conformational distributions of the three nucleoprotein segments didn’t change much,and they can partially fold at the isolated state,so this is a conformation selection process.While in the final binding process,the conformational distributions shift obviously to the folded state.In the final example we explored the T4 lysozyme conformational change with ligand binding mechanism.Proteins usually have more than one native structure.The local parts or domains move relatively to make the conformation of proteins change and fit for the ligand binding.T4 lysozymes have many native structures and a series of them are measured by experiments,and the conformations are found to transform,especially in the ligand binding process.We constructed the double-basin structure-based potential for the conformational dynamics of T4 lysozyme simulation,and build the backbone of the ligand peptidoglycan and the interactions between the two molecules to simulate the binding process coupled with the conformational change of T4 lysozyme.Our results show the conformational dynamics of T4 lysozymes,that it can jump between the domain open and closed state before ligand binding.After binding with ligand,the conformation state mainly stay at the domain closed state.T4 lysozyme has obvious N-and C-terminal domain,connected by the middle helix,and the conformational change is due to the domain motions.The binding mechanism with ligand can also separate with the two domain binding sequence with ligand.In the 2D free energy distribution of the contacts between N-terminal domain and the C-terminal domain with the ligand,two paths are shown with the binding process in N-terminal domain first or last.But no path through the partial binding of both N-and C-terminal domain path,means that at the early stage,T4 lysozyme N-and C-terminal domains cannot by constrained by the ligand.But after partial binding,the ligand can hold both N-and C-terminal domains and change the conformation distribution of T4 lysozyme.But single-molecule experiments report that T4 lysozyme will show some intermediate,and some stable structures are crystallized in experiments with the domain open angle between the most open and closed ones.These structures in solution perhaps resident with very short time,besides,our double-basin structure-based potential is designed for double-state conformational change,the ligand binding process may influence the conformational distribution of T4 lysozyme,but the ligand model seems too simple to simulate the real complicated interactions between T4 lysozyme and the ligand.In this work,molecular dynamics simulations show provide systematic explorations on protein coupled binding-folding,flexibility tuning in binding process and conformational change.The sampling methods with parallel methods can accelerate the statistics in the conformational space.Besides,coarse-grained techniques with structure-based force field provide high efficiency in the molecular simulations.The unfrustrated potential makes the protein conformational change and interaction process much faster.The replica exchange molecular dynamics provide the ability to sample large protein and complicated interactions.Our atomistic level simulations are all supported by this method.