To understand how the stabilities of key nuclei fragments affect protein folding dynamics, we simulate by molecular dynamics （MD） simulation in aqueous solution four fragments cut out of a protein G, including one a...To understand how the stabilities of key nuclei fragments affect protein folding dynamics, we simulate by molecular dynamics （MD） simulation in aqueous solution four fragments cut out of a protein G, including one a-helix （seqB： KVFKQYAN）, two -turns （seqA： LNGKTLKG and seqC： YDDATKTF）, and one -strand （seqD： DGEWTYDD）. The Markov State Model clustering method combined with the coarse-grained conformation letters method are employed to analyze the data sampled from 2-#s equilibrium MD simulation trajectories. We find that seqA and seqB have more stable structures than their native structures which become metastable when cut out of the protein structure. As expected, seqD alone is flexible and does not have a stable structure. Throughout our simulations, the native structure of seqC is stable but cannot be reached if starting from a structure other than the native one, implying a funnel-shape free energy landscape of seqC in aqueous solution. All the above results suggest that different nuclei have different formation dynamics during protein folding, which may have a major contribution to the hierarchy of protein folding dynamics.展开更多
基金Supported by the National Basic Research Program of China under Grant No.2013CB932804the National Natural Science Foundation of China under Grant No.11421063the CAS Biophysics Interdisciplinary Innovation Team Project
文摘To understand how the stabilities of key nuclei fragments affect protein folding dynamics, we simulate by molecular dynamics （MD） simulation in aqueous solution four fragments cut out of a protein G, including one a-helix （seqB： KVFKQYAN）, two -turns （seqA： LNGKTLKG and seqC： YDDATKTF）, and one -strand （seqD： DGEWTYDD）. The Markov State Model clustering method combined with the coarse-grained conformation letters method are employed to analyze the data sampled from 2-#s equilibrium MD simulation trajectories. We find that seqA and seqB have more stable structures than their native structures which become metastable when cut out of the protein structure. As expected, seqD alone is flexible and does not have a stable structure. Throughout our simulations, the native structure of seqC is stable but cannot be reached if starting from a structure other than the native one, implying a funnel-shape free energy landscape of seqC in aqueous solution. All the above results suggest that different nuclei have different formation dynamics during protein folding, which may have a major contribution to the hierarchy of protein folding dynamics.
