Grant: $608,782 - National Science Foundation - Jun. 6, 2009
25% voted satisfied - 75% voted not satisfied - 4 vote(s) cast
Award Description: Detailed information on protein dynamics at an atomic level and its thermodynamic implications is of fundamental biophysical importance for understanding protein stability and function. The only currently available tools to probe such information are nuclear magnetic resonance (NMR) and molecular dynamics (MD) computer simulation. They already have had enormous impact on our understanding of the energy landscapes of proteins and their complexes. Each of these approaches, however, has its own limitations. The goal of our research is the development of new methods in combining experimental and in silico tools for a comprehensive and accurate description of complex dynamics and thermodynamics of proteins, and, in addition, to demonstrate their applicability to biologically important molecular systems. Dynamics time scales that can be detected by NMR range from pico- to milliseconds and beyond. The advent of increasingly long MD computer simulations promise to close the time-scale gap between simulations and experimental NMR measurements, in particular residual dipolar couplings, chemical shifts, and rotating frame relaxation rates. Recently, long MD simulations into the microsecond range with the latest generation of molecular mechanics force fields have been conducted in our research group and the data generated were compared with experiment for the protein backbone as well as the side chains. In some cases, the agreement is remarkable where as in others there is room for improvement. These studies not only yield a comprehensive assessment of the accuracy of force fields and provide guidance for their further optimization, but they also provide a realistic microscopic approach to the interpretation of these experimental NMR parameters. The MD simulations also allow analysis of statistically significant correlated motions between soft degrees of freedom, such as dihedral angles, which are difficult to capture experimentally and thereby permit the accurate extraction of configurational entropies providing a direct quantitative link between dynamics and thermodynamics. By analyzing the motional correlations over time scales that cover many orders of magnitude, a better understanding of the statistical relevance and functional importance of these effects will be developed. These concepts are presently being tested in our laboratory on various protein systems, including ubiquitn and calbindin D9k. Furthermore, these concepts are applied to the experimental and computational analysis of the entropy transfer mechanism of the protein MDM2 when it interacts with the tumor-suppressor p53 or antagonistic inhibitors. The results of this research are expected to produce new computational, statistical thermodynamic, and NMR spectroscopic concepts for the characterization of proteins, which will enhance the understanding of protein behavior and function and serve as input for the engineering of proteins with new properties and the design of better drugs. Some of the tools developed will be made available to the biophysics and biomolecular NMR communities in terms of public web servers. The project provides cross-disciplinary training and research opportunities for graduate students and post-docs at the Department of Chemistry and Biochemistry at Florida State University (FSU), at the Institute of Molecular Biophysics at FSU, and at the National High Magnetic Field Laboratory (NHMFL).
Project Description: We have developed an all-atom local contact model for the prediction of protein motions underlying isotropic crystallographic B-factors. It uses a mean-field approximation to represent the motion of an atom in a harmonic potential generated by the surrounding atoms resting at their equilibrium positions. Based on a 400-ns molecular dynamics simulation of ubiquitin in explicit water, it is found that each surrounding atom stiffens the spring constant on average by a term that scales exponentially with the interatomic distance. This model combines features of the local density model and the local contact model. When applied to a non-redundant set of 98 very high-resolution protein structures an average Pearson correlation coefficient of 0.75 is obtained for all atoms. The systematic inclusion of crystal contact contributions and fraying effects is found to enhance the performance substantially. Because the computational cost of the local contact model scales linearly with the number of protein atoms, it is applicable to proteins of any size permitting the prediction of B-factors of both backbone and side-chain atoms. The model has similar or better performance than several other models tested, such as rigid-body motional models, the local density model, and various forms of the elastic network model. It is concluded that at the currently achievable level of accuracy collective intramolecular motions are not essential for the interpretation of B-factors. Since this contact-model based predictor is very rapid, it is possible to offer it as a public web server on our laboratory's homepage. It permits uploading of a protein databank (PDB) file, returns the predicted B-factors for all available atoms, including crystal contacts, and graphically displays the comparison between experimental and predicted B-factors. A beta version of this web server is now available on our web portal at http://spinportal.magnet.fsu.edu/bfactors/bfactors.html .
Jobs Summary: This project allowed us to retain two temporary support positions. (Total jobs reported: 1)
Project Status: Less Than 50% Completed
This award's data was last updated on Jun. 6, 2009. Help expand these official descriptions using the wiki below.