Understanding the relationship between enzyme structure and catalysis

Predicting the location of the active site, and the identity of the catalytic residues, is an important first step for annotating an enzyme of unknown function. We have developed a neural network trained to distinguish catalytic and non-catalytic residues based on a mixture of sequence and structural parameters. We find that the correct location of the active site can be predicted in ~70% of cases. We also find that the most important factor in making a prediction is the conservation score of each residue. However, including structural data does improve the predictions that are made over those made on the basis of conservation alone. Our first analysis of enzyme structure aims to measure the extent of conformational change undergone upon substrate binding. We find that most enzymes do not undergo large scale conformational change, and in many cases the catalytic residues are isolated from changes that do occur. One new theory of enzyme action suggests that certain, very small, conformational changes (deriving from changes in the enzyme dynamics) are important in catalysis. In a separate study, we look for these changes and find hints that they may occur in some enzymes, although they are at the limit of what can be reliably observed in crystal structures. Having established that catalytic site is generally preformed, we next look at the interactions between catalytic residues. We analyse a large set of interactions the roles they play in catalysis. We find that many catalytic residues do not require direct interactions with other catalytic residues to be active, although we do predict that many previously unidentified interactions will prove to be functionally important. Those residues that do commonly interact are often important in the pKa modulation of other residues.EThOS - Electronic Theses Online ServiceGBUnited Kingdo