Protein allostery. computational characterization of diguanylate cyclase PleD

In biology, accurate cellular regulation in response to environmental signals is crucial for the fitness of organisms. On the molecular level the modulation of protein activity is often achieved by the binding of a signaling molecule or by covalent modifications such as phosphorylation by kinases. Protein allostery, that is signal propagation from a distant allosteric site to functional sites to regulate the output, has long been recognized but the views and perspectives have been strongly influenced by different scientific fields and the continuous development of new methods. In particular, computational approaches are suited to bridge the gap between structure and dynamics and provide insight at an atomic level. After reviewing experimental and theoretical methods to study allostery, results from computational methods applied to the diguanylate cyclase PleD are presented. First, structural and dynamical aspects of the communication between the allosteric inhibition site and the active site are highlighted by energy calculations and molecular dynamics simulations. Ligand binding may trigger a balance-like movement of the conserved strand β2 that potentially displaces residues required for catalysis. In addition, dynamical coupling between the functional sites, i.e. simultaneous quenching of motion upon ligand binding, is found from normal mode analysis. Furthermore, two possible communication pathways connecting the inhibition with the active site are proposed. Second, processes involved in PleD dimerization were elucidated. In dynamics simulations the spontaneous active-to-inactive transition is observed and implies changes in the D1/D2 interface together with a slight decrease in the dimerization contact area. In the proposed model the β4-α4 loop repositioning is followed by adjustments in the α4-β5-α5 face that are amplified by the extended helix α5 by a leverage effect.