The Closed Open Transition of Adenylate Kinase From Crystal Structures and Computer Simulations

Many proteins function as dynamic molecular machines that cycle between well-defined states. A mechanistic and atomic-scale understanding starts with crystal, NMR or electron microscopy structures in these states. Typically, none or only very limited structural information is available for the intermediates along the transition. Computational methods can simulate transitions between states but due to the absence of intermediate structures it is hard to verify that the simulated transition path is correct. One exception is the enzyme adenylate kinase. It is well studied and a large number of crystal structures are available. Vonrhein et al [1] suggested early on that some of these structures would be transition intermediates due to stabilization by crystal contacts and created a `movie' from nine structures. We took this idea one step further and compare 45 experimental structures to hundreds of transitions of E. coli AdK simulated with the dynamic importance sampling method (DIMS). We find that DIMS trajectories, which only require a crystal structure for the starting and the end point of the transitions, contain all intermediate crystal structures (RMSD for matches: <4 A with median 1.2 A). The crystal structures can be time-ordered according to the DIMS trajectory and this ordering is essentially the same for forward and backward transitions. These results suggest that DIMS is capable of simulating realistic macromolecular transitions. From the simulated trajectories we can present a molecular detailed picture of a macromolecular transition. We discuss the conformational change of AdK with respect to the presence or absence of ligands, the relevance of salt bridges, and the motions of rigid domains.[1] C. Vonrhein, G. J. Schlauderer, and G. E. Schulz. Movie of the structural changes during a catalytic cycle of nucleoside monophosphate kinases. Structure 3 (1995),483–490.