Exploring the intermediate states of ADP-ATP exchange: a simulation study on Eg5.

While mitotic kinesins have attracted significant attention in recent years as new anticancer drug targets, the underlying mechanism of kinesin-catalyzed ATP hydrolysis is still under investigation. Crystal structures of Eg5, one of the best-studied kinesins, have been solved in both ADP-bound and ATP-bound states. However, it is still extremely challenging to experimentally obtain structural information on the functionally important intermediate states, such as the nucleotide free (apo) and the initial ATP-kinesin collision state. Systematic molecular dynamics simulations were performed in this study to mimic different nucleotide binding states and explore the critical structural and dynamic variations during ADP-ATP exchange. Clear conformational changes from "ADP-like" toward "ATP-like" were observed from the simulation results. A highly conserved residue Arg(234) was found to play a key role during the nucleotide exchange. This positively charged residue acted as the "hub" of a hydrogen-bond network that extended the effect of γ-phosphoryl group to both SW-I and SW-II regions. Comparison among the results of different nucleotide binding states indicated that the existence of γ-phosphoryl was immediately sensed at the initial ATP collision state by residue Ser(233), and this initial interaction induced the "back-door" opening and the "front-door" closing of the nucleotide binding pocket. In addition, several potential allosteric binding sites were identified through combination of correlation analysis and binding site mapping approaches based on the simulated apo ensemble, which provided additional targeting sites for novel allosteric Eg5 inhibition. These molecular simulation results provided not only a better understanding of Eg5-catalyzed ATP hydrolysis but also the structural basis for design of novel specific Eg5 inhibitors as anticancer therapeutic agents.