| Proteins consist of a sequence of residues. These residues generate the electrostatic nature of the protein, thereby determines all other properties for the protein. Computational techniques for calculating electrostatic properties are well established. More recently, bioinformatics techniques, used to investigate and compare protein sequences, have come into prominence. Combining these separate methodologies should have a synergistic effect. In this work, electrostatic based techniques will be developed and applied to simulate the flux of ions through membrane channels. A procedure for the prediction of ion flow, based on full three dimensional charge distribution a specific pH, and containing no freely adjusted variables, is presented. This procedure is used to gain important insight on the mechanism for the regulation of potassium ion flow in KcsA and provides support for a possible method of channel opening. Additionally, bioinformatic based techniques are developed and applied to the Cytochrome P450 and Fe2S2 Ferredoxin families. Insights gained from these techniques are integrated into electrostatic investigation of the protein-protein association of a member from each protein family, P450 cam and Putidaredoxin. These two proteins function as an electron-transfer pair and bind using charge complementary sites on their surface. The specific binding site is unknown. Three dimensional Brownian dynamics methods were developed and applied to identify possible binding sites. Many such interacting residue pairs were determined and evaluated. Several exhibited a high reaction probability. These sites were shown to demonstrate properties in agreement with experimental observations and thus are predicted to be good candidates for P450cam-Putidaredoxin binding sites. |
