Testing the role of transition state complementarity and the *enzyme environment in ketosteroid isomerase

Enzymes greatly accelerate chemical reactions required for biological life. Although much is understood about the mechanisms by which enzymes function, many questions remain. This thesis focuses on the role of the enzyme active site environment in stabilizing the transition state, and specifically on the contribution to catalysis of electrostatic and geometric complementarity to the transition state. Both electrostaic and geometric complementarity are suggested to be important for catalysis, but it has been difficult to distinguish these mechanisms or determine quantitative contributions to catalysis relative to an uncatalyzed reaction. I investigated these questions in ketosteroid isomerase (KSI), an enzyme suggested to use electrostatic complementarity in an oxyanion hole for catalysis. By systematically varying the charge localization in the oxyanion hole using a series of phenolate transition-state analogs, I found at most a 300-fold rate enhancement from electrostatic complementarity to the transition state. These results suggested an environment little better than water at stabilizing transition-state charge buildup, and this model was confirmed through the use of a sterically hindered series of phenolates that could not form optimal hydrogen bonding interactions at the oxyanion hole, and also supported through electrostatic calculations. I then examined the role of remote binding interactions with KSI's steroid substrate in excluding solvent from the active site and thereby changing the environment by comparing the reactivities of a full-length and truncated substrate. While there was a large catalytic contribution from remote binding interactions, there was no differential effect on oxyanion hole mutations, suggesting a similar environment with or without remote binding interactions. Several residues that contribute to catalysis through remote binding interactions were also identified. It seemed possible that oxyanion hole hydrogen bonds might be geometrically complementary to only the transition state oxyanion and not the ground state carbonyl oxygen; binding of ground state analogs was unaffected by oxyanion hole mutation and crystallographic analysis suggested poor hydrogen bonding geometry for the ground state. Finally, the oxyanion hole tyrosine was extensively mutated to both natural and unnatural residues, leading to an estimate of a 200-fold advantage from the enzymatic hydrogen bond relative to solution hydrogen bonding in the uncatalyzed reaction.