The geometry of metal sites is intimately connected to the electronic structure of a metal, and thus ultimately to metallo-protein function. Fully, one-third of all proteins in the Protein Databank, which are used extensively by researchers world-wide, are metalloproteins. However, in the majority of these structures the accuracy of the metal ligand bonds is too low to make statements about the metal properties as reflected in the geometry. With the advent of freezing techniques and synchrotron radiation it is now feasible to collect high (1.7 Angstroms_ to ultra- high resolution data (0.8 Angstroms) on many metalloproteins. At the same time increases in computer speed and memory sizes finally allow rigorous full-matrix analyses of this data to determine standard uncertainties (formerly known as estimated standard deviations) of the bond lengths and angles. Metalloproteins are especially amenable to such analyses because the greater proportional scattering power of the electron-dense metal atom results in an increase in the positional certainty, and thus an increase in the accuracy of bond lengths and bond angles including the metal. For metalloproteins with data of resolution less than about 1.7-1.5 Angstroms, it becomes possible to remove all restraints on the metal to ligand distances thus removing any bias such restraints might have on metal site geometry. We propose to determine a number of iron-metalloprotein structures, including ones from other project members, to 1.7-0.8 Angstroms resolutions, refine the structures with anisotropic thermal parameters, hydrogens and split side-chains (as data to parameter ratios allow), and analyze the structures with full- matrix least-squares to determine standard uncertainties to determine the geometry of other metal sites including copper, zinc, calcium and others. The resulting metal-ligand bond lengths and bond angles will be added to our metalloprotein database and distributed via the World Wide Web. These highly accurate structures with known standard uncertainties will help answer questions about the extent to which proteins can distort metal structures, and allow more accurate calculations of metal electronic structures, ultimately leading to better understanding of metalloprotein structure and function.
Showing the most recent 10 out of 53 publications
