W0221

Non-uniform Anion Distribution in Solvent Space of Cubic Insulin Crystals. Bin Yu¹ and Donald L.D. Caspar², ¹Ph.D Program 804 Uris Hall, Columbia University, NYC, NY 10027, ²Institute of Molecular Biophysics, Florida State University, Tallahasse, FL 32306 USA

From collection of complete low resolution data and iterative refinement of low resolution phases, there is strong evidence that anions at high ionic strength form multiple layers in the solvent channel.

Protein crystals contain large amount of solvent and significant volume in both the solvent and solute molecules do not have well-defined average positions. Mapping the distribution of solvent and solute can help us to gain important insights in the dynamics of the protein molecule inside the crystal. Most literature to date have only attempted to model the combined electron density in the solvent space, we were able to isolate the contribution of anions by isomorphous replacement. Phases of the low resolution data are refined by iterative density modification in real space with the high resolution structure as the primary constraint. Very complete X-ray data were collected in both high and low resolution on cubic insulin crystals at pH9 in high concentration of glucose and 1M salt (sulfate at room temperature and tungstate at LN). The protein atoms and ordered solvent molecules were refined with CCP4 and X-PLOR against high resolution data, and showed no significant change in the ordered structures. However, at lower resolution, we see both differences in the refined low resolution phases and statistically significant differences in observed structure factor to about 10 Å.

The difference in electron density between the crystals that comes from the tungstate distribution is represented by the vector difference map computed with observed structure factors and independently refined phases. RDF of the solvent difference map shows convincingly the existence of relatively broad maxima at 4.2 A and 8.5 A. Both the position and shape of the peaks are quite different from the "water shells" observed in crystals at low salt concentrations. This suggests that anions at high ionic strength form multiple shells around the protein surface.