Regulating transcription regulators via allostery and flexibility

Allosteric control of transcriptional regulatory proteins enables organisms to respond to changes in environmental and metabolic conditions. Eukaryotic and prokaryotic transcriptional regulatory proteins sense the availability of a broad range of small molecules including metal ions, metabolites, and drugs with the final biological outcome of altering transcriptional activity of specific genes. This alteration of transcriptional activity is a direct consequence of the ability of the small molecule to elicit a change in the affinity of the transcription factor for its target regulatory site on DNA. Consistent with allosteric mechanisms in general, our understanding of molecular mechanisms by which small-molecule binding perturbs DNA binding by transcriptional regulatory proteins is incomplete. The article by Reichheld et al. (1) in this issue of PNAS provides evidence that the allosteric mechanism in the tetracycline repressor (TetR) is centered on the ability of the small ligand, tetracycline, to alter the folding properties of the repressor protein.

The TetR senses intracellular tetracycline concentration. Antibiotic binding to TetR leads to a decrease in affinity of the protein for DNA that allows transcription of the genes that code for TetA, a membrane protein that exports the tetracyline out of the bacterial cell before it can attack its target, the ribosome, and TetR itself (2). The repressor protein is a homodimer in which each monomer is composed of a tetracycline binding/dimerization domain and a helix–turn–helix DNA binding domain (3). High-resolution structures of the DNA-bound apo (unliganded) and holo (tetracycline-bound) TetR indicate two distinct structures and form the basis of an allosteric model in which tetracycline binding shifts the protein from a conformation that is competent to bind DNA to one that is not (3–5). Reichheld et al.'s (1) protein folding studies provide support for an alternative model in which the DNA binding domain of the unliganded or high-affinity TetR samples an ensemble of conformations and folds independent of the tetracycline binding domain. Antibiotic binding leads to cooperative folding of the entire protein and “locks” the DNA binding domains into a conformation that has reduced affinity for DNA. The model has implications for our understanding of several phenomena including allostery, the role of disorder in binding, and the evolution of transcription regulatory proteins.

Models for allosteric regulation of transcriptional regulatory proteins owe much to the availability of high-resolution structures determined, in large part, by X-ray crystallography. In these models binding of a small molecule, either inducer or corepressor, alters the affinity of the transcription regulator for DNA by shifting the protein between low-affinity and high-affinity conformations. For example, small-molecule inducer binding can shift the DNA binding domain from a conformation that is complementary to the DNA target to one that is not structurally complementary to the target. The value of high-resolution structures to elucidation of the mechanisms of functional biochemistry is unquestionable. However, the incorporation of results of genetic studies and equilibrium and kinetic measurements of biochemical function along with structure provides an unparalleled strategy in developing self-consistent models of biomolecular function.

Allosteric mechanisms of several well-characterized transcriptional regulatory proteins, including the Escherichia coli tryptophan (6–8), and biotin repressors (9), and the E. coli catabolite repressor protein (10), involve some degree of ligand-induced folding. In each of these proteins the small molecule acts as a corepressor and its binding promotes DNA binding by promoting folding. Presumably the loss of flexibility accompanying effector binding freezes out conformations that are not productive for binding and/or lowers the entropic penalty for binding. By contrast, in the model for TetR allostery that emerges from the studies of Reichheld et al. (1) it is the unliganded, conformationally heterogeneous form of TetR that is characterized by high-affinity DNA binding (Fig. 1). Although the connection between folding and binding is well documented, the mechanism by which protein disorder can result in higher affinity is not known (11). Disorder is hypothesized to provide a kinetic advantage in binding by increasing the collisional cross-section available for binding (12). In this “fly-casting mechanism” the unstructured protein provides a large surface for initial ligand encounter. Once the encounter complex forms, the system is converted via a unimolecular process to a final folded ligand-bound state. Limited experimental data, primarily from NMR spectroscopy measurements, on the detailed mechanisms of binding processes characterized by disorder-to-order transitions are available. However, the limited data are consistent with the multistep process predicted by the fly-casting mechanism.

Fig. 1.

Model of allosteric regulation of a conformationally flexible transcription regulatory protein. The unliganded or apo protein exits as an ensemble of energetically equivalent structures. Small ligand and DNA binding drive the protein to distinct conformational preferences.

The significance of disorder in site-specific DNA binding has been reported. DNA binding at specific sites occurs through a multistep mechanism in which nonspecific binding facilitates the search for specific sites (13) and, consequently, the role of flexibility in both steps must be considered. In the ETS transcription factor flexibility is correlated with high-affinity binding to the target site on DNA and loss of this flexibility via phosphorylation results in decreased DNA binding affinity (14). The millisecond to microsecond time-scale motions observed in the lactose repressor DNA binding domain are thought to be important for nonspecific binding that is the prelude to facilitated diffusion of the protein to its specific binding site on DNA (15). Thus, the flexibility in the DNA binding domain of the apo or high-affinity form of TetR is not unprecedented. How this flexibility facilitates site-specific DNA binding by TetR remains to be determined.

The mechanism of allosteric communication in proteins, and the role of flexibility in transmission of allosteric signals, is currently the subject of much research. Fundamentally allostery is a thermodynamic phenomenon in which the binding of one molecule (or a posttranslational modification event) alters the affinity with which a protein binds to a second molecule (16, 17). In the classical two-state Monod–Wyman–Changeux model the structural basis for this functional behavior is attributed to the ability of the allosteric protein to adopt two alternative conformations (18). The allosteric effector, by virtue of its preference for one versus the other structure, drives the system toward one or the other conformation. Models for several allosteric transcriptional regulatory proteins derived from high-resolution structures are consistent with this two-state model. The inclusion of a disordered or a conformationally heterogeneous state (Fig. 1) in the mix does not preclude the application of a two-state model. Rather, instead of defining the system by two distinct structures, it is described by an ensemble of structures in one functional state and a much narrower ensemble in the other. Allosteric effector binding, either the small molecule or the DNA, simply limits the conformations that are energetically available to the protein (Fig. 1).

Disorder is hypothesized to provide a kinetic advantage in binding.

The TetR is a member of a large family of related proteins that is estimated, from sequence analysis, to contain >2,000 members. (19) These proteins respond to a broad range of signaling molecules and are involved in transcriptional regulatory processes that control, among other things, antibiotic biosynthesis, efflux pumps, and osmotic stress. Structures of these proteins thus far determined indicate a high level of conservation among family members. Assuming that the allosteric model proposed by Reichheld et al. (1) applies to even a small fraction of the TetR family one must considered what advantage the disorder associated with the DNA binding domain provides for the signaling properties of the protein. Based on statistical thermodynamic modeling, allosteric coupling is maximized when the domains containing one or both binding sites are intrinsically disordered a significant fraction of the time in the absence of ligand (20). Thus, the intrinsic disorder associated with the DNA binding domain of TetR may serve to maximize negative coupling between tetracycline and DNA binding. The tight coupling has the biological consequence of rendering the repressor a very effective relay of intracellular tetracycline concentration to gene expression, thereby permitting survival of the cell.