The role of dynamics in allosteric regulation

Abstract

The biomolecular conformational changes often associated with allostery are, by definition, dynamic processes. Recent publications have disclosed the role of pre-existing equilibria of conformational substates in this process. In addition, the role of dynamics as an entropic carrier of free energy of allostery has been investigated. Recent work thus shows that dynamics is pivotal to allostery, and that it constitutes much more than just the move from the ‘T’-state to the ‘R’-state. Emerging computational studies have described the actual pathways of allosteric change.

Introduction

Dynamic processes in biomolecules cover a large timescale regime, including very fast fluctuations of individual atoms on the picosecond timescale, loop and domain motions on the nanosecond timescale, conformational rearrangements on the millisecond timescale and breathing modes on a timescale slower than seconds. Several excellent reviews summarize the functional role of dynamics in catalysis 1., 2.. Here, we review the role of dynamics in allostery. Allostery (‘allo-steric = other-space’) means that action in one part of the molecule causes an effect at another site. Allosteric processes are closely associated with ligand-induced conformational changes that propagate between the allosterically coupled binding sites [3]. The allosteric molecule thus changes its coordinates as a function of time, which constitutes dynamics. Hence, dynamics and allosteric processes are almost tautologically linked. In addition, changes in the dynamic properties of the different conformations of the allosteric protein may contribute to the free energy of allosteric coupling through entropic effects [4]. This review focuses on the different types of dynamics of allosteric proteins. Purists may associate only quasi-harmonic motions on the shortest timescales with the word dynamics. In this review, however, we will consider all motions — quasi-harmonic, diffusive rearrangements or fluctuations in populations over an ensemble of subconformations — as dynamics. The review ignores the role of allosteric proteins in systemic dynamic processes, such as the dynamic changes of gene expression and neuron proliferation.

It is often stated that allosteric systems are oligomeric and symmetric (e.g. [5]); for this review, we want to take a broader point of view. We define here as allosteric those systems in which the binding of one ligand affects the affinity of another ligand (Figure 1). This includes, in addition to the classical homotropic oligomeric systems such as hemoglobin and aspartate transcarbamylase, heterotropic monomeric systems such as Hsp70 chaperones, whereby allosteric coupling exists between ATP binding in one domain and protein substrate binding in another domain [6], and single-domain proteins, whereby phosphorylation at one site affects the structure of the protein in a remote area 7., 8.••. Our definition also includes those proteins for which at least one of the ligands is a biological macromolecule; good examples are the Trp- and RNA-binding attenuation protein repressor, for which tryptophan binding affects RNA binding at a remote site [9^••], and processes such as ligand-coupled oligomerization 10., 11..

We want to mention a few developments that have made many of the exciting recent discoveries in allosteric dynamics feasible. Improvements in mutagenesis and site-specific labeling techniques extended the applicability of fluorescence lifetime methods beyond naturally available tryptophan residues [12]. Ultrafast laser technologies can resolve many time steps in dynamic processes [13]. The measurement of site-resolved quasi-harmonic dynamics on the picosecond to nanosecond timescale by NMR relaxation methods is feasible for both protein backbone and protein sidechains 14., 15., 16.. Fluctuations between conformational substates on the millisecond to microsecond timescale can now be measured quantitatively and site resolved by NMR as well [17]. Importantly, NMR dynamic measurements can now be combined with TROSY detection methods [18], which extends its application to larger systems, such as a 91 kDa allosteric protein [9^••]. The NMR measurement of residual dipolar couplings in solution has recently been applied to characterize dynamic process as well [19]. The commercial availability of precise microcalorimeters allows a quantitative assessment of entropic effects, which, in some cases, can be related to changes in dynamic processes [20]. A new, unsuspected source of dynamic information is ultra-high resolution low-temperature X-ray crystallography; anisotropic B-factors allow motional models for atomic and collective motions to be developed, and the observation of multiple conformations is suggestive of fluctuations over conformational substates [21^•]. Comprehensive synthetic approaches make it possible to design small allosteric systems [22^•]. Last but not least, the steady increase in computational power and the development of new computational approaches for the study of protein dynamics push simulations into time regimes that have previously not been accessible [23]. This permits the comparison of simulated and experimental dynamics, ultimately allowing predictions of allosteric dynamic pathways.

To the best of our knowledge, this subject has been reviewed systematically only once before [24]. That work is an excellent source for an overview of movements in allosteric proteins as inferred from static structures, NMR methods for the detection of picosecond to nanosecond backbone dynamics, and the dynamic properties of the allosteric Trp repressor system. In the following, we selected several representative examples from the enormous volume of literature that has been reported since then and have organized the information in four sections.