EVOLUTION OF A CATALYTICALLY EFFECTIVE MODEL ENZYME: THE IMPORTANCE OF TUNED CONFORMATIONAL FLUCTUATIONS

Possible causal connections between the dynamics of a thermally fluctuating model enzyme molecule and catalysis are explored. The model is motivated by observations from experiment and simulation that amino acid residues residing in different enzymatic domains may show markedly different degrees of conformational freedom. Consequently, we are interested in the catalytic efficacy of an enzyme as a function of long-range many-atom cooperative effects resulting from strong, moderate, and weak interactions between enzymatic residues. Here we show and quantify through molecular dynamics simulations how the number and distribution of these interactions affects an enzyme's conformational fluctuation dynamics and its effectiveness as a catalyst. For any given distribution of "stiff" and "loose" enzymatic domains, catalytic fitness is defined as the number of chemical events — specifically the number of times a catalytic residue and substrate surmount a chemical reaction barrier — during molecular dynamics simulation. Through mutation, recombination, and a selection procedure following the ideas of Darwinian evolution, a genetic algorithm drives a population of enzyme molecules to greater catalytic fitness by modifying the mix of stiff and loose interactions. Approximately 30,000 different enzyme molecules are generated by the genetic algorithm — each with a unique number and distribution of strong, moderate, and weak inter-residue interactions. While the catalytically least fit enzyme exhibits 16 chemical events, the fittest boasts 253. That point mutations far from the active-site chemistry in the fittest enzyme have a strong effect on the number of chemical events suggests that catalysis depends, in part, on long-range many-atom globally correlated dynamical fluctuations.