The Computational Biophysics Section studies problems of biological significance using several theoretical techniques: molecular dynamics, molecular mechanics, modeling, ab initio analysis of small molecule structure, and molecular graphics. These techniques are applied to a wide variety of macromolecular systems. Dr. Stan's research has focused on chaperonin-mediated protein folding. Protein folding mediated by chaperonin molecules is studied by computer simulations. Our focus is on the GroEL-GroES chaperonin complex of the Escherichia coli. Annealing action of the GroEL-GroES chaperonin system involves large scale conformational changes which the chaperonin system employs to prevent aggregation or misfolding of proteins. We analyze the structural transformations of GroEL during the cycle by examining the known structures and dynamics of GroEL Identifying natural substrates for chaperonins accomplished by using a sequence-based approach. The E. coli chaperonin machinery GroEL assists the folding of a number of proteins. We describe a sequence-based approach to identify the natural substrate proteins (SPs) for GroEL. Our method is based on the hypothesis that natural SPs are those that contain patterns of residues similar to those found in either GroES mobile loop and/or strongly binding peptide in complex with GroEL. The method is validated by comparing the predicted results with experimentally determined natural SPs for GroEL. We have searched for such patterns in five genomes. In the E. coli genome we identify 1422 (about a third) sequences that are putative natural SPs. A limited analysis of the predicted binding sequences shows that they do not adopt any preferred secondary structure. Our method also predicts the putative binding regions in the identified SPs. The results of our study show that a variety of SPs, associated with diverse functions, can interact with GroEL. Dr. Klauda's research consists of three areas; structural and dynamical behavior of lipids, membrane proteins, and beta-hairpin folding. As a first step in the lipid work, the aliphatic portion of C27 force field was improved (referred to as C27r) using high-level quantum mechanical calculations on alkanes. Molecular dynamics simulations of DPPC bilayers with C27r resulted in improved agreement with experiment compared to the C27 parameter set for gauche populations, NMR deuterium order parameters, and 13C NMR relaxation times. Collaborations with the experimental groups of Prof. John Nagle (Carnegie Mellon University) and Dr. Klaus Gawrisch (NIH/NIAAA) are underway to improve our understanding of structural and dynamical behaviors of lipid bilayers. In addition, the mechanism and protein structural response to sugar transport in lactose permease of E. coli., a membrane protein, is being studied with molecular dynamical simulations of the protein embedded in a POPE bilayer. The final part of Dr. Klauda?s work involves the study of beta-hairpin folding using the Self-Guided Langevin Dyanamics (SGLD) method recently developed. The folding pathways of several design peptides are being studied with explicit solvent and the efficiency of SGLD is being evaluated with explicit solvent. Dr. Zheng's research involves two projects based on the normal modes analysis of a highly simplified elastic network model: 1. We have developed novel computational methods to systematically analyze functionally relevant dynamic correlations within macromolecular complexes: we define two types of dynamic correlations (fluctuations-based and density-based), then we use them to select dynamically important ?hinge residues? and decompose the selected dynamical correlations down to individual normal modes to identify the most relevant modes. We have applied these methods to the analysis of the motor domain of dictyostelium myosin and have obtained interesting results that shed light on its mechanism of force generatioon. 2. We propose a statistical correlation for residues of protein complexes between their dynamical importance to a functionally relevant normal mode (measured by which is computed based on perturbational normal mode analysis of an elastic network model) and their sequential conservation obtained from multiple sequence alignment. We test this relation for a variety of DNA/RNA polymerases, and find it to hold for the open/closed conformational change that is common to them: the clusters of high-dw residues match well with the clusters of conserved residues for a significant portion of the sequence covering the fingers and the palm domains. The clusters of high-dw residues contain conserved residues that are not involved in substrate binding and may be conserved for their dynamic importance to the open/closed conformational change represented by the relevant normal mode. Dr. Che's research involves computer-aided molecular design, which applies principles of ab initio calculations and molecular dynamics simulations to the deign and development of bioactive agents such as drugs. The projects include successful design of following small molecules to act as protein secondary structure mimetics. For inhibiting protein-protein binding, general methods for mimicking protein surface structures would represent a significant advances. Protein binding interfaces comprise alpha-helix, beta-sheet, turns, polyproline structures, and loops secondary structures in an apparently unbiased manner, thus we have designed small molecules that can mimic each of these structures. Dr. Woodcock's research involves the investigation of carbohydrate structure, specifically the conformational properties of glucose analogs and examining the effects that solvation has on these systems. A collaboration with the group of Dr. Daron Freedberg and involves investigating the structural parameters of glucose (determined via NMR) and exploring the various effects that lead to the observed structural characteristics.
Showing the most recent 10 out of 45 publications
