News and Publications

Synthesis of Multiresolution Structural Data

W. Wriggers, P. Chacón, C.L. Brooks III, R.A. Milligan, E. Kim*

* University of California, Los Angeles, CA

Future advances in modern biology and medicine will depend on an understanding of fundamental cellular processes, most of which involve the actions and interactions of large biomolecular aggregates. Three-dimensional structures and image reconstructions of aggregates, involving hundreds of thousands to millions of atoms, are now routinely determined by using x-ray crystallography and electron microscopy. To further our understanding of cellular function, we must be able to characterize the architecture and functionally relevant motions of these cellular machines at all levels of resolution.

The Computational Structural Biology program is active both in developing computational tools for multiresolution docking and modeling and in applying this technology to the study of large-scale molecular machines, including molecular motors (myosin and kinesin) and their tracks (actin and microtubules), virus capsids, and the ribosome. We use topology-representing neural networks for a coarse estimation of density maps and for determining suitable landmarks for the registration of multiresolution data (Fig. 1). In the past year, we showed the usefulness of this approach in a variety of docking applications and developed the program package Situs to provide a powerful method for the localization of biomolecular subunits in low-resolution data from electron microscopy and small-angle x-ray scattering. Using neural networks to examine discrete parts of the search space reduced the search times for optimum alignment of 3-dimensional data from hours to seconds.

The rigid-body docking technique also laid the groundwork for the development of a flexible docking technique that brings deviating features of multiresolution structures into register. We devised a novel flexible docking procedure based on artificial neural networks and molecular dynamics simulation that improves the agreement between data sets in the presence of "induced fit" conformational changes. This approach was applied to the flexible docking of the ribosomal elongation factor EF-G into electron microscopy density maps.

Finally, we investigated how to include geometric constraints from a variety of biophysical sources in the modeling: electron microscopy, small-angle x-ray scattering, fluorescence spectroscopy, mutagenesis, and biochemical labeling and footprinting. For example, we refined the atomic model of actin filaments by using cysteine cross-links that were introduced experimentally and that provide stringent constraints for the arrangement of actin subunits and their contact interface. To investigate the effect of H73, conserved in all actin species, on protein structure, we complemented mutagenesis experiments with simulated annealing simulations of the mutants that revealed an effect of the electrostatic charge at this position on the forming of salt bridges near the active site.

PUBLICATIONS

Kim, E., Wriggers, W., Phillips, M., Kokabi, K., Rubenstein, P., Reisler, E. Cross-linking constraints on F-actin structure. J. Mol. Biol. 299:421, 2000.

Wriggers, W., Agrawal, R.K., Drew, D.L., McCammon, J.A., Frank, J. Domain movements of EF-G bound to the 70S ribosome: Insights from a hand-shaking between multi-resolution structures. Biophys. J. 79:1670, 2000.

Wriggers, W., Milligan, R.A., McCammon, J.A. Situs: A package for docking crystal structures into low-resolution maps from electron microscopy. J. Struct. Biol. 125:185, 1999.

Xing, J., Wriggers, W., Jefferson, G.M., Stein, R., Cheung, H.C., Rosenfeld, S.S. Kinesin has three nucleotide-dependent conformations: Implications for strain-dependent release. J. Biol. Chem., in press.

Yao, X., Grade, S., Wriggers, W., Rubenstein, P. His₇₃, often methylated, is an important structural determinant for actin: A mutagenic analysis of His₇₃ of yeast actin. J. Biol. Chem. 274:37443, 1999.