Allostery in Its Many Disguises: From Theory to Applications

Abstract

Allosteric regulation plays an important role in many biological processes, such as signal transduction, transcriptional regulation, and metabolism. Allostery is rooted in the fundamental physical properties of macromolecular systems, but its underlying mechanisms are still poorly understood. A collection of contributions to a recent interdisciplinary CECAM (Center Européen de Calcul Atomique et Moléculaire) workshop is used here to provide an overview of the progress and remaining limitations in the understanding of the mechanistic foundations of allostery gained from computational and experimental analyses of real protein systems and model systems. The main conceptual frameworks instrumental in driving the field are discussed. We illustrate the role of these frameworks in illuminating molecular mechanisms and explaining cellular processes, and describe some of their promising practical applications in engineering molecular sensors and informing drug design efforts.

Keywords: Allostery; allosteric drugs; allosteric material; allosteric switches; elastic network models; energy landscape; molecular dynamics; protein conformational changes; protein function; regulation; signal transduction.

Figure 1.. Mechanistic Underpinning of Allostery: Insights from Computations and Experiments

(A) Artist rendering of the conformational transitions network of the photoactive yellow protein, the 125-residue water-soluble blue-light photoreceptor from H. halophile, mapped onto the energy landscape of the system using the simulation procedures of Bolhuis and collaborators. (B) Frustration-based allostery in the human glucocorticoid receptor (GR), an intrinsically disordered transcription factor analyzed by Li et al. (2017). (I) Domain organization of the constitutively active GR constructs for translational isoforms, wherein the intrinsicallydisordered (ID) N-terminal domain (NTD) varies in length. Residues 1–97 (red) are labeled R (for Regulatory) and residues 98–420 (gray) are labeled F (for Functional). Also labeled are residues corresponding to the activation function 1 core (AF1 core) region, which is required for transcriptional activity. (II) Competing thermodynamic coupling in GR produces frustration. Schematic view ofthe thermodynamic configuration of GR. According to the displayed convention, the positive (+) signs between the DNA-binding domain (DBD) and F-domain, and the DBD and R-domain signify they are positively coupled; stabilization of one domain stabilizes the other. The negative (—) sign between the R- and the F-domains indicates that they are negatively coupled; stabilization of one domain destabilizes the other. (C) The closed conformation adenylate kinase observed upon ligand binding is sampled by the open form apo structure, illustrating the work of Bahar. (I) Two experimentally resolved structures, unbound (left) and ligand-bound (right). (II) Conformer predicted by ENM analysis to be accessible via a soft mode to the unbound structure. Blue and green refer to different domains. The substrate is shown in orange spheres (adapted from Temiz et al., 2004). (D) Distributions of GroEL molecules with different numbers of bound ATP molecules at different ATP concentrations from thework of Horovitz and co-workers. (E) Allosteric regulation in CRISPR-Cas9, by Palermo and McCammon. (I) Dynamical network model of CRISPR-Cas9, identifying groups (or “communities”) of closely correlated residues and the strength of correlation between them before (top) and upon (bottom) PAM binding. (II). The allosteric path between the spatially distance HNH and RuvC domains of the Cas9 protein.

Figure 2.. Allosteric Toy Models and Allosteric Materials

(A) Schematics of the allosteron model of McLeish, in binding (A) and self-assembly (B) illustrating local changes to spring constants κ, and the introduction of coupling springs between allosteron units κ_c. (B) Illustration of the work of Wyart and collaborators: (I) response (black) arrows to a stimulus (purple arrows) in random spring network decays rapidly with distance, i.e., there is little action at a distance. (II) Networks can be evolved in which there is specific action at a distance. Note that the response is amplified near the active site (blue arrows), indicating the presence of a lever in the structure. (III) Example of hinge architecture obtained while optimizing cooperativity, in which two parts of the material rotate around a hinge located at the center of the system. (IV—VI) Illustration of the cooperative architectures found: hinge (clothespin), shear (mint box), and twist (Rubik’s cube).

Figure 3.. Rational Design of Allosteric Systems and Identification of Allosteric Sites

(A) Schematic diagrams illustrating the work of Dokholyan and colleagues on optogenetlc and chemogenetic control of target proteins using allostery and protein order-disorder transition (reprinted from Dagliyan et al. 2016). (B) Illustration of the approach by involving the chemical rescue of the active conformation of a protein. The example shows how mutation of a buried tryptophan to glycine leads to a structural disruption—either through a discrete conformational change or through loss of protein stability—that leads to loss of protein function. Adding exogenous indole can then complement the cavity caused by the deleted side chain, restoring the original protein conformation and, thus, its function. (C) Principle of the rational design and engineering of a synthetic DNA-based nanodevice described by Plaxco. Top: the designed cooperative DNA-nanodevice comprises the recognition element consisting of a triplex-forming DNA sequence, which behaves like a “clamp” that binds a specific 9-base DNA ligand via the formation of both Watson-Crick and Hoogsteen base-pair interactions. The cooperative DNA-nanodevice is obtained by joining together two sequential copies of one-half of such recognition element linked via a flexible 22-base, single-stranded loop (gray portion) to two sequential copies of its other half. Binding of the ligand to the first receptor decreases the entropic cost associated with the binding to the second receptor (and thus improves its affinity for the ligand). As a result, this nanodevice shows a Hill-type cooperative response, with a Hill coefficient n_H = 2.1 ± 0.1 (figure reproduced from Mariottini et al. 2017). (D) Binding hotspots of small chemical probes to flexible regions of the protein tend to correspond to cryptic binding sites. Example from the work of Kozakov, showing the mapping of hotspots identified by FTsite in the unbound structure of the catalytic subunit of thecAMP-dependent protein kinase A (PDB: 2GFC, chain A) displayed in tan. Three hotspots, obtained after domain splitting, are shown as clusters of molecular probes: a cluster of 18 probes (cyan); a cluster of 16 probes (magenta); a cluster of 13 probes (gray). An inhibitor (yellow) is superimposed for reference.