Enzymes are the most versatile catalysts. Because of their exquisite selectivity, diverse array of catalyzed reactions, mild reaction conditions, and significant enhancement of reactions rates, enzymes isolated from natural sources have been widely used in the chemical, biotechnology and health care industries. However, unfavorable intrinsic properties of enzymes, including marginal stability, narrow substrate specificity and incompatibility with nonaqueous solvents, have made engineering of enzymes necessary. For some applications, enzymes can be designed de novo to catalyze reactions that are not found in nature. Therefore, our abilities to design and redesign efficient enzymes have extremely important and practical implications. The rational redesign of enzymes towards increased catalytic activity and stability is an ultimate test of our understanding of protein sequence-structure-function relationships. Although advances in computational tools have enabled construction of new enzymes catalyzing unnatural reactions, our ability to drastically improve enzyme activity towards a desired reaction in a rational manner has remained underdeveloped. In contrast, directed evolution experiments, in which fitness is elevated via random mutation and selection, is a highly successful method of improving enzyme function. However, our understanding of the structural and mechanistic basis of beneficial random mutations and fitness landscape remains rudimentary. Therefore, a comprehensive investigation of how large sequence changes can lead to dramatic changes in enzyme function will not only bridge this fundamental knowledge gap in protein sequence-structure-function relationships, it will also significantly improve our capabilities in designing custom enzymes with desired properties. This proposal attempts to reveal the structural and mechanistic bases of protein fitness landscape by combining the expertise of a protein engineering lab (Yi Tang), a structural biology lab (Todd Yeates) and a computational protein design lab (Ken Houk). Four interrelated and interdisciplinary aims will explore the catalytic landscape of a recently discovered enzyme, LovD, whose activity has been newly evolved in the laboratory towards the synthesis of the cholesterol lowering drug simvastatin: 1) Structural analysis of mutants in the directed- evolutionary pathway of a LovD enzyme carrying out a new reaction; 2) Biochemical and biophysical studies of LovD and mutants; 3) Computational characterization of LovD mutants; and 4) Computational prediction of alternate sequence mutations expected to confer enhanced catalytic activity on LovD.
Understanding the sequence-structure-function relationships of proteins is one of the most important goals in biological sciences. We proposed to perform comprehensive structural and mechanistic analysis of a set of evolved mutants of the simvastatin synthase LovD obtained from an exhaustive directed evolutionary effort. By using both experimental and computational tools, we aim to provide insights into enzyme catalysis and to enable more intelligent rational (re)design of new enzymes.
