Showing papers by "David A. Case published in 1997"

Domain packing and dynamics in the DNA complex of the N-terminal zinc fingers of TFIIIA.

Abstract: The three N-terminal zinc fingers of transcription factor IIIA bind in the DNA major groove. Substantial packing interfaces are formed between adjacent fingers, the linkers lose their intrinsic flexibility upon DNA binding, and several lysine side chains implicated in DNA recognition are dynamically disordered.

Incorporating solvation effects into density functional theory: Calculation of absolute acidities

Abstract: An approach to the calculation of molecular electronic structures, solvation energies, and pKa values in condensed phases is described. The electronic structure of the solute is described by density functional quantum mechanics, and electrostatic features of environmental effects are modeled through external charge distributions and continuum dielectrics. The reaction potential produced by a mode of the molecular charge distribution is computed via finite-difference solutions to the Poisson-Boltzmann equation and incorporated into the self-consistent field procedure. Here we report results on three sets of organic acids, whose pKa values range over 16 pH units. The first set provides models for ionizable side chains in proteins; the second set considers the effects of substituting one to three chlorine atoms for hydrogens in acetic acid; and the final set consists of 4-substituted-bicyclo-[2.2.2]-octanecarboxylic acids. Successful prediction of “absolute” pKa values places stringent requirements on the computation of gas-phase proton affinities and on the response to solvation. In some cases the current model shows substantial errors, but overall the results and trends are in good agreement with experiment. Prospects for extending this approach to more complex systems such as proteins are briefly discussed. © 1997 John Wiley & Sons, Inc.

Density Functional Calculations of Proton Chemical Shifts in Model Peptides

Abstract: Density-functional chemical shielding calculations are reported for the alanine dipeptide with a variety of backbone torsion angles and for methane and N-methylacetamide complexes with rare gases, monatomic ions, water, and other amides. These fragment systems model electrostatic, nonbonded, and hydrogen bonding interactions in proteins and have been investigated at a variety of geometries. The results are compared to empirical formulas that relate intermolecular shielding effects to peptide group magnetic anisotropies, electrostatic polarization of the C -H and N-H bonds, magnetic contributions from C-C and C-H bonds, and close contact effects. Close contacts are found to deshield protons involved in close nonbonded contacts that typically occur in hydrogen bonds. "Lone pair" charges improve the model for electrostatic effects and are important for understanding the angular dependence of shifts for protons involved in hydrogen bonds. C-C and C-H bond anisotropy contributions help to explain the torsional dependence of amide proton shifts in alanine dipeptide. Good agreement is found between the empirical formulas and the quantum chemistry results, allowing a reassessment of empirical formulas that are used in the analysis of chemical shift dispersion in proteins.

Density Functional Study on the Electronic Structures of Model Peroxidase Compounds I and II

Abstract: The electronic structures of [Fe(Por)(Im)O]1+ and [Fe(Por)(Im)O] (model compounds I and II, respectively) have been studied on the basis of density functional theory or DFT (Por = porphine, Im = imidazole). The a2u π-cation radical state (4A2u) was determined to be the ground state of compound I with total spin equal to 3/2, while the a1u π-cation state (4A1u) was found to be 0.15 eV higher in energy than the 4A2u state. Since, in both states, the spins were localized to the porphyrin ring (S = 1/2) and the Fe−O center (S = 1), the magnetic coupling interaction between the two spin sites was examined by using a broken symmetry method. The calculated J value revealed very weak magnetic coupling for the A2u state, which corresponded to the experimental data. The calculated J value revealed strong antiferromagnetic coupling for the A1u state. The calculated Mossbauer spectrum parameters (quadrupole splitting and asymmetry) were similar for both the A1u and A2u states, and both agreed well with experimental v...

A computational study of the role of solvation effects in reverse turn formation in the tetrapeptides APGD and APGN

Abstract: The tetrapeptides APGD and APGN are known by NMR analysis to adopt reverse turn conformations to a significant degree in aqueous solution. We have carried out a 7.7 ns molecular dynamics simulation...

Normal mode analysis of biomolecular dynamics

Abstract: The past decade has seen an impressive advance in the application of molecular simulation methods to problems in chemistry and biochemistry. As computer hardware has become faster and software environments more sophisticated, the amount of detailed information available and its expected level of accuracy has grown steadily [1]. But it has become increasingly clear that the ‘easy’ part of a simulation project is setting up and carrying out the calculations, and the hard part generally lies in extracting useful data from among the very many things that can be calculated from a trajectory or Monte Carlo simulation. Normal mode analysis provides an approximate but analytical description of the dynamics, and has long been recognized as an important limiting case for molecular dynamics in condensed phases [2]. A principal limitation arises from the fact that normal modes are defined by an expansion about a particular point on the potential energy surface, and hence have difficulty describing transitions from one local minimum to another. The quasiharmonic and ‘instantaneous’ mode theories discussed below attempt to ameliorate some of this neglect of the ‘rugged’ nature of protein energy landscapes. Yet there remains a ‘paradoxical aspect’ [3] of biomolecular dynamics that is still the subject of considerable study: even though the energy surface contains many local minima, proteins behave in some ways as though the energy surface were harmonic, and normal mode analyses are often more correct than one might expect. This article reviews some recent experience on the application of normal mode ideas to biomolecules, looking at how this technique describes short-timescale motion as well as longer timescale, collective motions. The use of harmonic ideas in the analysis of crystallographic and NMR data will also be outlined.

Quantum Mechanical Modeling of Active Sites in Metalloproteins. Electrostatic Coupling to the Protein/Solvent Environment

Abstract: Metalloproteins are quite common in biochemistry: more than 50% of the structures in the Protein Data Bank, and an estimated one-third of all purified enzymes, require a metal ion for their biological function [1]. The study of methods to treat such systems in simulations is very active. In this chapter, we will discuss some approaches to energetic issues involving transition metal clusters in proteins, drawing examples primarily from recent work in our lab. We will focus on two issues. The first involves the use of density functional theory, and the extent to which it can provide accurate and reliable results for the first row transition metals. A second area will discuss our experience, and that of others, in using continuum solvent models as a framework for constructing descriptions of the interaction of metalloprotein active sites with their protein and solvent environments. We will consider primarily energetic and thermodynamic issues, with an emphasis on energetic effects of charge changes, such as adding protons or electrons to a system. This is a fairly “static” picture that leaves out important kinetic and dynamic effects; nevertheless, establishing sound information about thermodynamic “end points” is a key (and often very difficult) step in constructing models for the function of metalloproteins.